Comparative Transcriptome Profiling Analysis of Red- and White-Fleshed Strawberry (Fragaria×ananassa) Provides New Insight into the Regulation of the Anthocyanin Pathway

Comparative Transcriptome Profiling Analysis of Red- and White-Fleshed Strawberry... Abstract Anthocyanins are water-soluble pigments in plants. They confer both economic and healthy profits for humans. To gain a deeper insight into the regulation of anthocyanin biosynthesis in octoploid strawberry (Fragaria×ananassa; Fa), a widely consumed economically important fruit, we performed comparative transcriptomic analysis of red- and white-fleshed strawberry cultivars in two ripening stages. In total, 365,455 non-redundant transcripts were assembled from the RNA sequencing (RNAseq) data. Of this collection, 377 were annotated as putative anthocyanin-related transcripts. Differential expression analysis revealed that 57 anthocyanin biosynthesis transcripts were down-regulated, and 89 transcription factors (TFs) were either down- or up-regulated under anthocyanin deficiency. Additionally, amongst the 50,601 putative long non-coding RNAs (lncRNAs) identified here, 68 lncRNAs were differentially expressed and co-expressed with differentially expressed anthocyanin-related mRNAs; 2,070 co-expressing lncRNA–mRNA pairs were generated. Expression profile analysis revealed that it was the limited expression of FaF3'H (flavonoid 3'-hydroxylase) that blocked the cyanidin 3-glucoside accumulation in the two investigated strawberry cultivars. This was further supported by a transient overexpression experiment with FaMYB10. The down-regulated lncRNAs might participate in anthocyanin regulation by acting as targets for microRNAs (miRNAs). The level of competitive intensity in miRNA and lncRNA for the same mRNA targets was probably lower in the white-fleshed strawberries, which can release the repression effect of the mRNAs in red-fleshed strawberry as a result. This study for the first time presents lncRNAs related to anthocyanins in strawberries, provides new insights into the anthocyanin regulatory network and also lays the foundation for identifying new anthocyanin regulators in strawberry. Introduction Strawberry (Fragaria×ananassa; Fa) is widely consumed not only for its enriched bioactive compounds but also for its attractive fruit color. These key quality traits are attributed to the compounds such as anthocyanins which are recognized as one of the most important antioxidants, thereby contributing to the healthful attributes (Afrin et al. 2016). Different types and content of anthocyanins bring us different colored strawberry cultivars ranging from orange to extremely dark red (Pillet and Folta 2015). Moreover, white, yellow, peach and pink-blushed strawberries also exist, suggesting distinct anthocyanin metabolism among cultivars. It is necessary to investigate the underlying metabolism in strawberries in order to improve fruit quality. To this end, the accessions with different colors, especially those deficient in anthocyanin accumulation, provide us with good opportunities. Anthocyanins are water-soluble pigments, belonging to the flavonoid class. They are glycosides or acylglycosides of polyhydroxyl, derivatives of 2-phenylbenzopyrylium or flavylium salts. Nearly 700 different anthocyanins have been identified so far (Andersen and Jordheim 2010). In strawberry, numerous anthocyanins have been identified, including 3-rutinosides of pelargonidin and cyanidin (Da Silva et al. 2007), pelargonidin 3-(malonyl) glucoside and pelargonidin 3-(6-acetyl)-glucoside (Wu and Prior 2005), cyanidin 3-(succinoyl) glucoside and pelargonidin 3-(succinoyl) glucoside (Wang et al. 2002). The pelargonidin 3-glucoside and cyanidin 3-glucoside, conferring bright and dark red color, respectively, have been recognized as two major anthocyanins (Sondheimer and Karash 1956, Mazza and Miniati 1993), while pelargonidin 3-glucoside occurs in a much higher amount than cyanidin 3-glucoside. It accounts for >70% of total anthocyanins (Wang et al. 2002, Kosar et al. 2004, Donno et al. 2013). Anthocyanins are derived from the branched flavonoid biosynthetic pathway. It is proven that the accumulation of anthocyanins is regulated by the expression levels of biosynthetic genes, including chalcone synthase (CHS), chalcone isomerase (CHI), flavanone 3-hydroxylase (F3H), flavonoid 3'-hydroxylase (F3'H), dihydroflavonol 4-reductase (DFR) and anthocyanidin synthase (ANS). (Salvatierra et al. 2010). If any step of these enzymatic catabolites is up-regulated or blocked/perturbed this can lead to variations in the final products, and then in the visual color. For instance, overexpression of F3'H from apple in Arabidopsis and tobacco enhanced the accumulation of anthocyanins and resulted in red seedlings and flowers, respectively (Han et al. 2010), while silencing of DFR (Lin et al. 2013) or F3H (Jiang et al. 2013) in strawberry has resulted in complete loss of red color in fruits. Besides these enzymatic reaction-related structural genes, transcription factors (TFs) are proved to be essential regulators in anthocyanin biosynthesis. Among them, MYB, basic helix–loop–helix (bHLH) and WD40-repeat (WD) proteins are the most extensively investigated types. They act independently or co-operate with each other as a ternary MYB–bHLH–WD40 (MBW) complex (Baudry et al. 2004, Xu et al. 2015) to regulate anthocyanin accumulation. In strawberry, several members of the MBW complex have been identified and characterized as regulators controlling proanthocyanidin biosynthesis (Schaart et al. 2013). What is more, knock-down of FvMYB10 resulted in undetectable concentrations of anthocyanins in woodland strawberry (Fragaria vesca; Fv), while the FvMYB10 overexpression lines had significantly elevated the anthocyanin levels (Lin-Wang et al. 2014), suggesting the pivotal role of MYB10 on regulating anthocyanin accumulation. This is also supported by the study describing a lack of anthocyanin production in strawberry when FaMYB10 was transiently silenced (Medina-Puche et al. 2014). On the other hand, knock-down of FvbHLH33 in woodland strawberry showed no effect on anthocyanin accumulation (Lin-Wang et al. 2014). In addition, repressors of anthocyanins were also identified, such as FaMYB1 (Aharoni et al. 2001), FaMYB5 and FabHLH3Δ (Schaart et al. 2013). More recently, many other TF families have been demonstrated to be involved in anthocyanin modulation, such as basic leucine zipper (bZIP) (An et al. 2017) and WRKY (Duan et al. 2018), suggesting the existence of complex regulatory networks. However, information about the regulation of anthocyanins by such TFs in strawberry is limited; more novel TFs need to be identified and characterized. Long non-coding RNAs (lncRNAs) are a class of functional RNAs with a length longer than 200 nt, lacking protein-coding capacity. They were originally thought to be transcriptional ‘noise’, due to their low expression and high sequence conservation compared with protein-coding mRNAs (Chekanova 2015, Liu et al. 2015a, Liu et al. 2015b, Shafiq et al. 2016). However, more and more evidence has been found to indicate that lncRNAs play critical regulatory roles in diverse biological processes in plants, including stress response (Wang et al. 2017), flower and fruit development (Zhu et al. 2015) and ripening (Kang and Liu 2015). In the past decades, numerous studies have been carried out to explore the functions of lncRNAs. It has been suggested that lncRNAs can regulate gene expression in both the cis- and trans-acting mode, exerting their functions on neighboring genes on the same allele, or distant alleles far away from where they were transcribed (Li and Rana 2012). Moreover, lncRNAs can interact with micro RNAs (miRNAs) by serving either as precursors to generate miRNAs, or as targets of miRNAs to compete for the miRNA binding with mRNA targets (Wang et al. 2017). As another type of non-coding RNAs (ncRNAs), some miRNAs have been reported to regulate the flavonoid and anthocyanin biosynthesis pathways (Gupta et al. 2017). For example, miR829.1 and miR1873 have been computationally identified in Podophyllum heandrum to target mRNAs coding for CHS and DFR, respectively (Biswas et al. 2016). Taken together, these results suggested that lncRNAs might participate in regulating anthocyanin biosynthesis. Although lncRNAs have been identified from several species, such as wheat (Cagirici et al. 2017) and pigeon pea (Nithin et al. 2017), information about lncRNAs and their expression profile response to anthocyanin deficiency in strawberry is still lacking. Through deep sequencing, the overall picture of regulation of the anthocyanin biosynthesis pathway can be investigated. To unravel the complex network of regulation of anthocyanin biosynthesis in strawberry, we generated transcriptomic profiling of anthocyanin-related genes in red-fleshed strawberry and its natural anthocyanin-deficient mutant (white-fleshed strawberries). Comparative transcriptomic analysis was conducted to assess why anthocyanins are deficient in the flesh of white-fleshed strawberries. Moreover, we carried out transcriptome-scale identification of lncRNAs in strawberry. The expression profiles of lncRNAs in response to anthocyanin deficiency were also analyzed. In addition, the potential regulatory network of lncRNAs–miRNAs–mRNAs was constructed, which gave us important clues regarding the roles of lncRNAs in regulating anthocyanin biosynthesis, providing new insights into anthocyanin regulation. Results Anthocyanin accumulation during strawberry fruit ripening We performed HPLC analysis of the major anthocyanins (cyanidin 3-glucoside and pelargonidin 3-glucoside) in the two strawberry cultivars (Fig. 1A). Strawberry ‘Benihoppe’ has red skin and red flesh; it is the progeny resulting from a cross between cv. ‘Akihime’ and cv. ‘Sachinoka’ (Mochizuki et al. 2014). ‘Xiaobai’ differs from ‘Benihoppe’ during tissue culture selection; it has red skin but white flesh. Since the two cultivars have similarly colored skin and different colored flesh, we examined the anthocyanins in skin (outer red layer including achenes) and flesh separately. As expected, anthocyanins were detected in the skin and flesh of ‘Benihoppe’ with a higher level in the fruit skin than in the flesh (Table 1). Anthocyanins were also detected in the skin of ‘Xiaobai’. On the other hand, no anthocyanins were detected in the white flesh of ‘Xiaobai’ during the whole fruit ripening process, from green (G) stage to the full red (R) stage (Table 1), which implied that anthocyanin biosynthesis has been blocked in the flesh of the cultivar as we can see from the phenotype (Fig. 1A). Further, the types of accumulated anthocyanins were analyzed. Interestingly, even in the red flesh of cv. ‘Benihoppe’, no cyanidin 3-glucoside accumulation was detected (Fig. 1B). Pelargonidin 3-glucoside was detected as the major anthocyanin type (Fig. 1B). However, in the skin, both ‘Benihoppe’ and ‘Xiaobai’ produced cyanidin 3-glucoside (Table 1, Fig. 1B), indicating different anthocyanin metabolism in strawberry skin and flesh. Table 1 Anthocyanin content (μg g-1 FW)a during strawberry fruit ripening ‘Benihoppe’ skin ‘Benihoppe’ flesh ‘Xiaobai’ skin ‘Xiaobai’ flesh Stages Cy Pg Cy Pg Cy Pg Cy Pg Green ND ND ND ND ND ND ND ND White ND ND ND ND ND ND ND ND Turning 30.4 ± 2.0 75.7 ± 18.5 ND 53.4 ± 10.2 26.2 ± 0.6 45.2 ± 5.2 ND ND Half red 44.0 ± 6.4 187.4 ± 21.1 ND 89.0 ± 6.1 32.2 ± 3.1 91.0 ± 4.1 ND ND Full red 43.8 ± 4.0 325.2 ± 37.6 ND 123.0 ± 15.1 37.6 ± 2.4 125.1 ± 6.0 ND ND ‘Benihoppe’ skin ‘Benihoppe’ flesh ‘Xiaobai’ skin ‘Xiaobai’ flesh Stages Cy Pg Cy Pg Cy Pg Cy Pg Green ND ND ND ND ND ND ND ND White ND ND ND ND ND ND ND ND Turning 30.4 ± 2.0 75.7 ± 18.5 ND 53.4 ± 10.2 26.2 ± 0.6 45.2 ± 5.2 ND ND Half red 44.0 ± 6.4 187.4 ± 21.1 ND 89.0 ± 6.1 32.2 ± 3.1 91.0 ± 4.1 ND ND Full red 43.8 ± 4.0 325.2 ± 37.6 ND 123.0 ± 15.1 37.6 ± 2.4 125.1 ± 6.0 ND ND a Anthocyanin content is represented as the average value ± SD of three biological replicates. Cy, cyanidin 3-glucoside; Pg, pelargonidin 3-glucoside; ND, not detected. Table 1 Anthocyanin content (μg g-1 FW)a during strawberry fruit ripening ‘Benihoppe’ skin ‘Benihoppe’ flesh ‘Xiaobai’ skin ‘Xiaobai’ flesh Stages Cy Pg Cy Pg Cy Pg Cy Pg Green ND ND ND ND ND ND ND ND White ND ND ND ND ND ND ND ND Turning 30.4 ± 2.0 75.7 ± 18.5 ND 53.4 ± 10.2 26.2 ± 0.6 45.2 ± 5.2 ND ND Half red 44.0 ± 6.4 187.4 ± 21.1 ND 89.0 ± 6.1 32.2 ± 3.1 91.0 ± 4.1 ND ND Full red 43.8 ± 4.0 325.2 ± 37.6 ND 123.0 ± 15.1 37.6 ± 2.4 125.1 ± 6.0 ND ND ‘Benihoppe’ skin ‘Benihoppe’ flesh ‘Xiaobai’ skin ‘Xiaobai’ flesh Stages Cy Pg Cy Pg Cy Pg Cy Pg Green ND ND ND ND ND ND ND ND White ND ND ND ND ND ND ND ND Turning 30.4 ± 2.0 75.7 ± 18.5 ND 53.4 ± 10.2 26.2 ± 0.6 45.2 ± 5.2 ND ND Half red 44.0 ± 6.4 187.4 ± 21.1 ND 89.0 ± 6.1 32.2 ± 3.1 91.0 ± 4.1 ND ND Full red 43.8 ± 4.0 325.2 ± 37.6 ND 123.0 ± 15.1 37.6 ± 2.4 125.1 ± 6.0 ND ND a Anthocyanin content is represented as the average value ± SD of three biological replicates. Cy, cyanidin 3-glucoside; Pg, pelargonidin 3-glucoside; ND, not detected. Fig. 1 View largeDownload slide Phenotypes and HPLC analysis of anthocyanins in red- and white-fleshed strawberries. (A) Phenotypes of red-fleshed strawberry ‘Benihoppe’ (left) and white-fleshed strawberry ‘Xiaobai’ (right). Black arrows indicate the consistent position where the flesh was collected for further experiments. (B) HPLC analysis of anthocyanins in the skin and flesh of red- and white-fleshed strawberries. Fig. 1 View largeDownload slide Phenotypes and HPLC analysis of anthocyanins in red- and white-fleshed strawberries. (A) Phenotypes of red-fleshed strawberry ‘Benihoppe’ (left) and white-fleshed strawberry ‘Xiaobai’ (right). Black arrows indicate the consistent position where the flesh was collected for further experiments. (B) HPLC analysis of anthocyanins in the skin and flesh of red- and white-fleshed strawberries. RNAseq and assembly The flesh of fruits at the white (W) stage and R stage of each cultivar with three biological replicates was used to build 12 libraries for high-throughput RNA sequencing (RNAseq). After removing the low-quality reads, around 100,000,000 clean reads of each sample were obtained. Overall, 55% of these reads could be mapped to the Fa strawberry reference genome (Supplementary Table S1). The unmapped reads with the corresponding mates were extracted and then de novo assembled. Finally, we obtained 365,455 non-redundant transcripts in total, with an average length of 643 bp, which were used for further analysis. Expression profiling of anthocyanin biosynthesis genes To investigate in which step anthocyanin synthesis was blocked in the white flesh of ‘Xiaobai’, we identified the candidate transcripts for anthocyanin biosynthesis by searching the standard gene names and mapping all transcripts to reference pathways in the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (Supplementary Fig. S1). In total, 377 transcripts involved in three pigment-related pathways, i.e. flavonoid biosynthesis, anthocyanin biosynthesis, and flavone and flavonol biosynthesis pathways, were identified (Table 2). The homologs in Fv are listed in Supplementary Table S2. Secondly, the abundances of these anthocyanin biosynthesis transcripts were estimated (Supplementary Table S2) and compared in white flesh and red flesh of strawberries. The transcripts with per kilobase million (TPM) value <2 across all samples were filtered out. As expected, all of the involved transcripts showed significantly down-regulated patterns in the white flesh of ‘Xiaobai’ compared with the red flesh of ‘Benihoppe’ strawberries. The only exception was observed in FaF3'H and two FaFLS transcripts, which showed no change or up-regulation in expression, respectively (Supplementary Table S3). The normalized abundances of differentially expressed transcripts were shown by the heatmap in Fig. 2. Moreover, a large proportion of these transcripts were inhibited at both the W stage and R stage in the white flesh of ‘Xiaobai’, while some were only down-regulated at either the W stage or R stage (Table 2; Supplementary Table S3). For instance, the expression of two transcripts encoding CHS (MSTRG.75989.1) and F3H (MSTRG.78844.1) was inhibited at both the W and R stage in the white flesh compared with the red flesh. Another CHS transcript (MSTRG.54516.2) was only down-regulated in the W stage, and one CHI transcript (TRINITY_DN43706_c1_g1_i1) was only down-regulated in the R stage. Table 2 Candidate genes related to anthocyanins in strawberry Pathways Genes Enzymes KO ids No. alla No. upb No. downc W R W R Anthocyanin biosynthesis CHS Chalcone synthase K00660 38 0 0 11 9 CHI Chalcone isomerase K01859 24 0 0 11 6 F3H Flavanone 3-hydroxylase K00475 20 0 0 13 13 F3'H Flavonoid 3'-hydroxylase K05280 6 0 0 0 0 DFR Dihydroflavonol 4-reductase K13082 61 0 0 9 4 ANS Anthocyanidin synthase K05277 17 0 0 7 4 UFGT Anthocyanidin 3-O-glucosylthransferase K12930 51 0 0 8 7 Anthocyanin modification GT1 Anthocyanidin 5,3-O-glucosyltransferase K12938 9 0 0 0 1 3GGT anthocyanidin 3-O-glucoside 2''-O-glucosyltransferase K12939 6 0 0 0 0 AA5GT Cyanidin 3-O-glucoside 5-O-glucosyltransferase (acyl-glucose) K17194 2 0 0 0 0 FAOMT Flavonoid 3',5'-methyltransferase K13272 3 0 0 0 0 5MaT1 Anthocyanin 5-O-glucoside-6''-O-melonytransferase K12934 1 0 0 0 0 UF3GT Flavonoid 3-O-glucosyltransferase K17193 86 0 0 1 0 Flavone and flavonol FLS Flavonol synthase K05278 16 0 2 0 4 Flavonone biosynthesis ANR Anthocyanidin reductase K08695 30 0 0 0 0 LAR Leucoanthocyanidin reductase K13081 7 0 0 0 0 Pathways Genes Enzymes KO ids No. alla No. upb No. downc W R W R Anthocyanin biosynthesis CHS Chalcone synthase K00660 38 0 0 11 9 CHI Chalcone isomerase K01859 24 0 0 11 6 F3H Flavanone 3-hydroxylase K00475 20 0 0 13 13 F3'H Flavonoid 3'-hydroxylase K05280 6 0 0 0 0 DFR Dihydroflavonol 4-reductase K13082 61 0 0 9 4 ANS Anthocyanidin synthase K05277 17 0 0 7 4 UFGT Anthocyanidin 3-O-glucosylthransferase K12930 51 0 0 8 7 Anthocyanin modification GT1 Anthocyanidin 5,3-O-glucosyltransferase K12938 9 0 0 0 1 3GGT anthocyanidin 3-O-glucoside 2''-O-glucosyltransferase K12939 6 0 0 0 0 AA5GT Cyanidin 3-O-glucoside 5-O-glucosyltransferase (acyl-glucose) K17194 2 0 0 0 0 FAOMT Flavonoid 3',5'-methyltransferase K13272 3 0 0 0 0 5MaT1 Anthocyanin 5-O-glucoside-6''-O-melonytransferase K12934 1 0 0 0 0 UF3GT Flavonoid 3-O-glucosyltransferase K17193 86 0 0 1 0 Flavone and flavonol FLS Flavonol synthase K05278 16 0 2 0 4 Flavonone biosynthesis ANR Anthocyanidin reductase K08695 30 0 0 0 0 LAR Leucoanthocyanidin reductase K13081 7 0 0 0 0 a No. all: the total number of transcripts that were analyzed. b No. up: the number of transcripts that were significantly up-regulated (|log2 FCe| |log2 FC| >1, FDR <0.05) in white-fleshed ‘Xiaobai’ compared with red-fleshed ‘Benihoppe’. c No. down: the number of transcripts that were significantly down-regulated (|log2 FC| >1, FDR <0.05) in white-fleshed ‘Xiaobai’ compared with red-fleshed ‘Benihoppe’. W, white stage; R, full red stage. Table 2 Candidate genes related to anthocyanins in strawberry Pathways Genes Enzymes KO ids No. alla No. upb No. downc W R W R Anthocyanin biosynthesis CHS Chalcone synthase K00660 38 0 0 11 9 CHI Chalcone isomerase K01859 24 0 0 11 6 F3H Flavanone 3-hydroxylase K00475 20 0 0 13 13 F3'H Flavonoid 3'-hydroxylase K05280 6 0 0 0 0 DFR Dihydroflavonol 4-reductase K13082 61 0 0 9 4 ANS Anthocyanidin synthase K05277 17 0 0 7 4 UFGT Anthocyanidin 3-O-glucosylthransferase K12930 51 0 0 8 7 Anthocyanin modification GT1 Anthocyanidin 5,3-O-glucosyltransferase K12938 9 0 0 0 1 3GGT anthocyanidin 3-O-glucoside 2''-O-glucosyltransferase K12939 6 0 0 0 0 AA5GT Cyanidin 3-O-glucoside 5-O-glucosyltransferase (acyl-glucose) K17194 2 0 0 0 0 FAOMT Flavonoid 3',5'-methyltransferase K13272 3 0 0 0 0 5MaT1 Anthocyanin 5-O-glucoside-6''-O-melonytransferase K12934 1 0 0 0 0 UF3GT Flavonoid 3-O-glucosyltransferase K17193 86 0 0 1 0 Flavone and flavonol FLS Flavonol synthase K05278 16 0 2 0 4 Flavonone biosynthesis ANR Anthocyanidin reductase K08695 30 0 0 0 0 LAR Leucoanthocyanidin reductase K13081 7 0 0 0 0 Pathways Genes Enzymes KO ids No. alla No. upb No. downc W R W R Anthocyanin biosynthesis CHS Chalcone synthase K00660 38 0 0 11 9 CHI Chalcone isomerase K01859 24 0 0 11 6 F3H Flavanone 3-hydroxylase K00475 20 0 0 13 13 F3'H Flavonoid 3'-hydroxylase K05280 6 0 0 0 0 DFR Dihydroflavonol 4-reductase K13082 61 0 0 9 4 ANS Anthocyanidin synthase K05277 17 0 0 7 4 UFGT Anthocyanidin 3-O-glucosylthransferase K12930 51 0 0 8 7 Anthocyanin modification GT1 Anthocyanidin 5,3-O-glucosyltransferase K12938 9 0 0 0 1 3GGT anthocyanidin 3-O-glucoside 2''-O-glucosyltransferase K12939 6 0 0 0 0 AA5GT Cyanidin 3-O-glucoside 5-O-glucosyltransferase (acyl-glucose) K17194 2 0 0 0 0 FAOMT Flavonoid 3',5'-methyltransferase K13272 3 0 0 0 0 5MaT1 Anthocyanin 5-O-glucoside-6''-O-melonytransferase K12934 1 0 0 0 0 UF3GT Flavonoid 3-O-glucosyltransferase K17193 86 0 0 1 0 Flavone and flavonol FLS Flavonol synthase K05278 16 0 2 0 4 Flavonone biosynthesis ANR Anthocyanidin reductase K08695 30 0 0 0 0 LAR Leucoanthocyanidin reductase K13081 7 0 0 0 0 a No. all: the total number of transcripts that were analyzed. b No. up: the number of transcripts that were significantly up-regulated (|log2 FCe| |log2 FC| >1, FDR <0.05) in white-fleshed ‘Xiaobai’ compared with red-fleshed ‘Benihoppe’. c No. down: the number of transcripts that were significantly down-regulated (|log2 FC| >1, FDR <0.05) in white-fleshed ‘Xiaobai’ compared with red-fleshed ‘Benihoppe’. W, white stage; R, full red stage. Fig. 2 View largeDownload slide Schematic of the anthocyanin biosynthesis pathway in strawberry. The expression patterns of transcripts were represented by the log2 TPM value and shown as a heatmap at the side of each step. The four cells from left to right represent the white stage of the red-fleshed cv. (W-R), the red stage of the red-fleshed cv. (R-R), the white stage of the white-fleshed cv. (W-W) and the red stage of the white-fleshed cv. (R-W), respectively. PAL, phenylammonia-lyase; C4H, cinnamate-4-hydroxylase; 4CL, 4-coumaroyl-CoA synthase; CHS, chalcone synthase; CHI, chalcone isomerase; F3H, flavanone 3-hydroxylase; F3′H, flavonoid 3′-hydroxylase; ANS, anthocyanidin synthase; FLS, flavonol synthase; DFR, dihydroflavonol 4-reductase; UFGT, anthocyanidin 3-O-glucosyltransferase; LAR, leucoanthocyanidin reductase; ANR, anthocyanidin reductase. Fig. 2 View largeDownload slide Schematic of the anthocyanin biosynthesis pathway in strawberry. The expression patterns of transcripts were represented by the log2 TPM value and shown as a heatmap at the side of each step. The four cells from left to right represent the white stage of the red-fleshed cv. (W-R), the red stage of the red-fleshed cv. (R-R), the white stage of the white-fleshed cv. (W-W) and the red stage of the white-fleshed cv. (R-W), respectively. PAL, phenylammonia-lyase; C4H, cinnamate-4-hydroxylase; 4CL, 4-coumaroyl-CoA synthase; CHS, chalcone synthase; CHI, chalcone isomerase; F3H, flavanone 3-hydroxylase; F3′H, flavonoid 3′-hydroxylase; ANS, anthocyanidin synthase; FLS, flavonol synthase; DFR, dihydroflavonol 4-reductase; UFGT, anthocyanidin 3-O-glucosyltransferase; LAR, leucoanthocyanidin reductase; ANR, anthocyanidin reductase. White flesh was recovered to red by overexpression of FaMYB10 To discover in depth the molecular reasons for losing color in the white flesh of strawberry ‘Xiaobai’, we successfully turned the white flesh back into red flesh by transient overexpression of one FaMYB10 TF gene (Fig. 3A). Cyanidin 3-glucoside was detected at a higher level in the entire FaMYB10-overexpressing ‘Xiaobai’ fruits (including both skin and flesh) than in the control (fruits infiltrated with bacteria containing empty vector) (Fig. 3B;Supplementary Fig. S2). This was different from what we have described in the previous section, i.e. that no cyanidin 3-glucoside was detected either in the red flesh of ‘Benihoppe’ or in the white flesh of ‘Xiaobai’ fruits (Fig. 1), indicating that FaMYB10 recovered the cyanidin 3-glucoside biosynthesis in strawberry. This proposal was also supported by overexpression of FaMYB10 in ‘Benihoppe’, as we also detected a higher concentration of cyanidin 3-glucoside in the flesh of FaMYB10-overexpressing‘Benihoppe’ fruits (Fig. 3B;Supplementary Fig. S2). To elucidate this matter, the expression levels of some key genes for anthocyanin biosynthesis in the red flesh of FaMYB10-overexpressing ‘Xiaobai’ fruits were estimated by real-time quantitative PCR (qPCR). As shown in Fig. 4, anthocyanin biosynthesis transcripts including FaDFR, FaANS and anthocyanidin 3-O-glucosylthransferase (FaUFGT) accumulated at a much higher level in the flesh of FaMYB10-overexpressing ‘Xiaobai’ fruits compared with the control (Fig. 4). Specially, the FaF3'H gene was expressed in the flesh of FaMYB10-overexpressing fruits, while it was rarely expressed in the normal red flesh of ‘Benihoppe’ and the white flesh of ‘Xiaobai’ fruits (Fig. 2; Supplementary Table S2). These findings revealed the key reason why cyanidin 3-glucoside accumulated in the flesh of FaMYB10-overexpressing fruits. Also, expression levels of TFs related to anthocyanin biosynthesis including WD and bHLH33 were estimated in the FaMYB10-overexpressing flesh. An increase in WD expression but no change in bHLH33 expression was observed, indicating a potential regulatory role for FaMYB10 in WD expression. Fig. 3 View largeDownload slide Phenotypes and anthocyanin content of FaMYB10-overexpressing strawberry. (A) The phenotypes of FaMYB10-overexpressing ‘Benihoppe’ (up) and ‘Xiaobai’ (down). White arrows indicate the position where the flesh color was changed after infiltration. (B) Anthocyanin content in FaMYB10-overexpressing fruits. 35s, control fruits infiltrated with Agrobacterium containing an empty vector; 35s:FaMYB10, experimental fruits infiltrated with Agrobacterium containing the recombined FaMYB10 overexpression plasmid. Fig. 3 View largeDownload slide Phenotypes and anthocyanin content of FaMYB10-overexpressing strawberry. (A) The phenotypes of FaMYB10-overexpressing ‘Benihoppe’ (up) and ‘Xiaobai’ (down). White arrows indicate the position where the flesh color was changed after infiltration. (B) Anthocyanin content in FaMYB10-overexpressing fruits. 35s, control fruits infiltrated with Agrobacterium containing an empty vector; 35s:FaMYB10, experimental fruits infiltrated with Agrobacterium containing the recombined FaMYB10 overexpression plasmid. Fig. 4 View largeDownload slide The expression of anthocyanin-related genes detected by qPCR in strawberry flesh. Bars represent the average relative transcript expression; error bars represent the SD of three biological replicates. R-W, the red stage of the white-fleshed cv. ‘Xiaobai’; R-R, the red stage of the red-fleshed cv. ‘Benihoppe’; 35s, control fruits infiltrated with Agrobacterium containing an empty vector; 35s:FaMYB10, experimental fruits infiltrated with Agrobacterium containing recombined FaMYB10 overexpression plasmid. Fig. 4 View largeDownload slide The expression of anthocyanin-related genes detected by qPCR in strawberry flesh. Bars represent the average relative transcript expression; error bars represent the SD of three biological replicates. R-W, the red stage of the white-fleshed cv. ‘Xiaobai’; R-R, the red stage of the red-fleshed cv. ‘Benihoppe’; 35s, control fruits infiltrated with Agrobacterium containing an empty vector; 35s:FaMYB10, experimental fruits infiltrated with Agrobacterium containing recombined FaMYB10 overexpression plasmid. qPCR validation of RNAseq To verify the gene expression data obtained by RNAseq analysis, eight transcripts involved in the anthocyanin biosynthesis pathway were selected for qPCR using the same RNA samples as used for sequencing (Supplementary Fig. S3). The results showed that despite the quantitative differences in expression levels, the expression patterns detected by qPCR experiments of most transcripts were consistent with the expression patterns investigated by RNAseq analysis. A linear correlation between the qPCR and the RNAseq data was observed (R2 = 0.73) (Fig. 5), indicating the reliability of the data. Fig. 5 View largeDownload slide Correlation analysis of qPCR relative expression and RNAseq fold change. Fig. 5 View largeDownload slide Correlation analysis of qPCR relative expression and RNAseq fold change. Identification of anthocyanin-related TFs As the structural genes involved in anthocyanin biosynthesis are largely regulated at the transcriptional level, we identified 89 TFs either significantly up- or down-regulated in W and R stages of white flesh compared with the red flesh strawberries (Supplementary Table S4). This collection included one MYB, three bHLH and three WD TFs. Intriguingly, a transcript (TRINITY_DN47718_c0_g1_i1) annotated as a single-repeat R3 type MYB (MYB1R) was much more highly expressed in the W stage of white-fleshed ‘Xiaobai’, with a log2 fold change (FC) (white/red) value of 8.6. In addition, a TTG1-like WD transcript, FaAN11 (TRINITY_DN42135_c0_g1_i3), exhibited significant down-regulation in the white flesh of ‘Xiaobai’ compared with the red flesh of ‘Benihoppe’, with a log2 FC (white/red) value of –5.1 in the R stage. In additon, many other TF types including ethylene-responsive factors (ERFs), WRKYs and NAC domain-containing proteins were also included in the differentially expressed TF sets (Supplementary Table S4), indicating the possibility of novel TFs participating in anthocyanin regulation. Characteristics of putative lncRNAs and their expression under anthocyanin deficiency Putative lncRNAs were identified based on our transcriptome data. First, among the 365,445 non-redundant transcripts, 299,935 transcripts with class_code ‘u’, ‘o’, ‘x’ and ‘i’ were extracted, and subsequently 241,307 transcripts passed the defined length criteria. A total of 54,244 transcripts were identified as non-coding transcripts after filtering by the coding potential calculator 2 (CPC2) server, and through the PLEK software. Potential protein domain-containing transcripts among these transcripts were removed by blasting against the Fv protein, Pfam and nr databases. The remaining 51,446 transcripts were kept for further analysis. To eliminate any known small RNAs, these transcripts were blasted against the tRNA database, rRNA database and miRNA database. Finally, 13.8% (50,601 transcripts) of the transcriptome assemblies were identified as putative lncRNAs (Supplementary Table S5). The length of lncRNAs ranged from 201 to 2,585 nt. The average length was 381 nt, and the average GC content was around 40%, with the highest GC content detected in shorter lncRNAs. A total of 41,570 transcripts with a TPM value >2 in at least three samples were identified as actively expressed putative lncRNAs (Supplementary Table S6). Differential expression analysis showed that when compared with red flesh from ‘Benihoppe’, 19 putative lncRNAs were significantly down-regulated, 12 were up-regulated at the W stage, and 49 and 20 lncRNAs were down-regulated or up-regulated at the R stage, respectively (Supplementary Table S7), in the white flesh of ‘Xiaobai’. Six and four transcripts were down- or up-regulated at both the R and W stage, respectively. The largest log2 FC (white/red) value was shown by TRINITY_DN1328_c1_g1_i1 (–10.1 at the R stage) and TRINITY_DN23526_c0_g1_i2 (13.29 at the W stage). Co-expression of lncRNAs and anthocyanin-related coding transcripts or TFs We carried out Pearson relative correlation analysis to investigate the trans-co-expression of differentially expressed lncRNAs, coding transcripts and TFs involved in the anthocyanin biosynthesis pathway. As a result, 64 lncRNAs interacted with 74 coding transcripts and TFs, establishing 2,070 lncRNA–mRNA pairs that were differentially co-expressed between white flesh and red flesh at a significant level (Supplementary Table S8). Most down-regulated lncRNAs showed a positive correlation, while up-regulated lncRNAs negatively correlated with mRNAs, except for FLS transcripts. Further, we observed that a particular lncRNA could be co-expressed with multiple coding transcripts and TFs, and multiple lncRNAs have been co-expressed with one particular coding transcript or TF. TRINITY_DN50364_c2_g1_i3 and TRINITY_DN47222_c0_g2_i7 appeared to be the most co-expressed lncRNA, being down-regulated and up-regulated, respectively. Construction of lncRNA–miRNA–mRNA networks Literature records have revealed that lncRNA functions are associated with miRNAs by acting as either their precursors or their targets. To investigate the details of this, 1,584 published miRNAs from strawberry were collected for network analysis. As a result, 130 lncRNAs were identified as putative precursors of 80 miRNAs (Supplementary Table S9). Interestingly, only one (TRINITY_DN50364_c2_g1_i3) of the precursor lncRNAs exhibited a down-regulated expression pattern in white-fleshed ‘Xiaobai’. It was detected as the precursor of miRNA PC-5p-142026_25, which also targeted one lncRNA transcript, TRINITY_DN49666_c1_g2_i21, showing no change in expression between the white- and red-fleshed cultivars. Moreover, lots of lncRNAs were identified as putative precursors of miRNAs which were previously known as regulators of flavonoid pathways, such as miRNA858a, miRNA156 and miRNA396b. However, no expression changes were detected in these lncRNAs. In order to provide more insights into lncRNA–miRNA association, the lncRNAs targets of miRNAs were predicted. The results showed that 60 differentially expressed lncRNAs were targeted by 392 miRNAs, establishing 591 anthocyanin-responsive miRNA–lncRNA target pairs (Supplementary Table S10). Further, the differentially expressed lncRNAs and their miRNA–mRNA network related to anthocyanins was visualized (Fig. 6). Among the interaction networks pairs, several miRNAs targeted one lncRNA and one mRNA, such as PC-5p-977833_4, fan-miR478, fan-miR2630, PC-5p-469001_8 and PC-5p-439332_9. In contrast, fan-miR2630, PC-3p-228442_16, PC-3p-217435_17, PC-5p-56438_63 and S4-m0037 targeted one lncRNA and several mRNAs. LncRNAs and mRNAs targeted by the same miRNAs were down-regulated in the white flesh of ‘Xiaobai’. Specifically, the up-regulated lncRNA TRINITY_DN44094_c0_g1_i15, TRINITY_DN1328_c0_g1_i1 and TRINITY_DN50080_c2_g3_i1 were targeted by miRNA PC-5p-2988_1087 and fan-miR869, respectively. Both miRNAs have one or more down-regulated mRNA targets. The down-regulated lncRNA TRINITY_DN43354_c0_g4_i1 was targeted by the miRNA PC-5p-150653_24, whose anthocyanin-related mRNA targets have been up-regulated. All these results indicated the complex interaction of lncRNAs, miRNAs and mRNAs. Fig. 6 View largeDownload slide Network between differentially expressed lncRNAs and anthocyanin-related protein-coding transcripts regulated by miRNAs. miRNA nodes were represented as triangles; lncRNAs and coding transcripts were represented as rectangles and circles, respectively. Transcripts that were up-regulated in white-fleshed ‘Xiaobai’ were colored red, and down-regulated transcripts were colored black. Fig. 6 View largeDownload slide Network between differentially expressed lncRNAs and anthocyanin-related protein-coding transcripts regulated by miRNAs. miRNA nodes were represented as triangles; lncRNAs and coding transcripts were represented as rectangles and circles, respectively. Transcripts that were up-regulated in white-fleshed ‘Xiaobai’ were colored red, and down-regulated transcripts were colored black. Discussion The role of FaF3'H in strawberry cyanidin 3-glucoside biosynthesis Anthocyanins are glycosides or acylglycosides of polyhydroxyl. Developmental programming plays a vital role in anthocyanin biosynthesis in strawberry. It was even thought to be the predominant factor over genotypes and environmental factors (Carbone et al. 2009). In support of this, our results showed that anthocyanin accumulation started from the T stage (the stage at which fruits start becoming red) and gradually increased with fruit ripening. The full red fruits had the highest level of anthocyanins (Table 1). In addition, anthocyanin accumulation exhibits tissue-specific characteristics. As previously suggested (Guan et al. 2016), the regulation of anthocyanin biosynthesis by light in the skin was different from that in the flesh of white-fleshed and teinturier grape berries. This might be caused by the differential expression of anthocyanin biosynthesis genes in different tissues. As an example, the anthocyanin biosynthesis genes (F3'5'H and ANS) were found to be differentially expressed in the skin and flesh of red-fleshed grapes (Xie et al. 2015). Additionally, in our results, cyanidin 3-glucoside was detected in the skin of both red-fleshed and white-fleshed strawberries, whereas in the flesh, no cyanidin 3-glucoside was detected. This finding was different from the previous reports describing that cyanidin 3-glucoside was detected at a considerable level in strawberry (Sondheimer and Karash 1956, Mazza and Miniati 1993, Lopes-da-Silva et al. 2002, Da Silva et al. 2007). This difference could be attributed to the differences in varieties. However, we detected cyanidin 3-glucoside accumulation in the flesh of FaMYB10-overexpressing fruits, suggesting that FaMYB10 has recovered the biosynthesis of cyanidin 3-glucoside. To elucidate in depth, we focused on the cyanidin biosynthesis branch. The two genes involved in this branch, ANS and UFGT, were expressed at a much higher level in the FaMYB10-overexpressing flesh compared with the control (Fig. 4). This difference might contribute to the cyanidin synthesis in FaMYB10-overexpressing fruits. According to this scenario, the red-fleshed ‘Benihoppe’ fruits with higher expression of these genes than the white-fleshed ‘Xiaobai’ should form cyanidin derivatives. However, even though these genes were more highly expressed in red-fleshed ‘Benihoppe’ than in white-fleshed ‘Xiaobai’ with the highest FC (red/white) of 7 (UFGT), the cyanidin derivatives are still absent in the flesh of red ‘Benihoppe’ fruits. Yet, it is not a satisfactory explanation for the lack of cyanidin derivatives in the white flesh of ‘Xiaobai’ strawberries. Therefore, F3'H, a necessary gene for formatting cyanidin, was considered. As generally known, F3'H catalyzes the first step in this branch. It was proven that a mutation in the coding region of the F3'H gene resulted in its inability to produce cyanidin-derived anthocyanins in morning glory (Zufall and Rausher 2004). Suppression of F3'H by RNA interference (RNAi) in chrysanthemum successfully shifted the predominantly cyanidin-derived pigments to non-pigment transgenic lines (Huang et al. 2013). On the other hand, overexpression of F3'H from apple in Arabidopsis resulted in more anthocyanin accumulation in mutants than in the wild type (Han et al. 2010). These results indicated the vital role of F3'H in anthocyanin accumulation. Our results showed that FaF3'H was rarely expressed in the red flesh of ‘Benihoppe’ and white flesh of ‘Xiaobai’, both of which had a lack of cyanidin derivatives. Hoever, it was expressed at a much higher level in FaMYB10-overexpressing flesh (Fig. 4), accompanied by a high level of cyanidin 3-glucoside accumulation. It is reasonable to conclude that the low expression of FaF3'H blocked the cyanidin 3-glucoside synthesis in the two studieded strawberry cultivars. The roles of anthocyanin biosynthesis genes in regulating anthocyanin accumulation Anthocyanin accumulation corresponds to the expression of genes encoding enzymes involved in the biosynthesis pathway. It has been previously suggested that major genes involved in anthocyanin biosynthesis were inhibited in a yellow strawberry variety at the turning stage (Zhang et al. 2015). Our study here showed that almost all key genes in anthocyanin biosynthesis were down-regulated from the W stage, indicating that the anthocyanin biosynthesis transcripts are perturbed at early developmental stages. This finding is consistent with conclusions that gene expression perturbation of flavonoid pathway genes in a white stawberry variety occurs in the early green developmental stage (Härtl et al. 2017). In our results, the expression levels of early biosynthetic genes such as CHS and F3H have been severely inhibited by 2- to 4-fold. The ‘late’ genes in anthocyanin biosynthesis, including DFR, ANS and UFGT, were also suppressed, with the largest log2 FC (red/white) around 7 in white-fleshed ‘Xioabai’. CHS catalyzes the first reaction step in forming the primary precursor for anthocyanin biosynthesis. If the expression of CHS was strongly inhibited, the anthocyanin concentration in strawberry fruits was significantly reduced, which finally led to the loss of fruit pigment accompanied by an increasing lignin content (Clark and Verwoerd 2011). F3H converts flavanone into dihydroflavanol. The absolute depletion of red color in F3H-silenced strawberry fruits revealed the critical role of F3H in blocking the anthocyanin biosynthesis (Jiang et al. 2013). DFR catalyzes the conversion of dihydroflavonol into colorless leucoanthocyanidins. Suppression of DFR in strawberry resulted in reduction of fruit red pigment (Lin et al. 2013), while overexpression of DFR could rescue the dfr mutant phenotype in Arabidopsis (Shin et al. 2016). ANS affects anthocyanin accumulation directly by converting the colorless leucoanthocyanidins to colored anthocyanidins (Reddy et al. 2007). However, we tried transient overexpression of one FaANS gene (accession: AY695817) in white-fleshed ‘Xiaobai’ strawberries; no pigment accumulated in the FaANS-overexpressing fruits (data not shown), suggesting that the lower expression of this structural gene might not be responsible for the loss-of-color ‘Xiaobai’ phenotype. However, we could not exclude its contribution to modulating anthocyanin content. In addition, UFGT catalyzes the key step for anthocyanin stability and water solubility in plants. The expression of UFGT was detected at a much higher level in red-fleshed than in white-fleshed cultivars, which is similar to the results observed in grape skin (Wu et al. 2017), indicating the role of UFGT in regulating anthocyanins. The regulation by FaMYB10 of anthocyanin pathways It has been reported that ANS is not regulated by transient silencing of FaMYB10 in Fa (Medina-Puche et al. 2014). Similar results have been found in RNAi FaMYB10 knocked-down Fv fruits (Lin-Wang et al. 2014). In contrast, we found that the expression levels of ANS were increased in FaMYB10-overexpressing flesh (Fig. 4). This finding is similar to the previous results showing that in the FaMYB10 overexpression lines, the expression of ANS was significantly up-regulated (Lin-Wang et al. 2014). This indicated the complex regulatory role of FaMYB10 in ANS expression. On the other hand, the expression of transcripts encoding TF FabHLH33 exhibited no change in FaMYB10-overexpressing fruits, indicating that FaMYB10 could not regulate the expression of FabHLH33 in Fa strawberry. This was supported by the fact that the expression of FvbHLH33 did not change either in FvMYB10-overexpressing or in FvMYB10-silenced Fv fruits (Lin-Wang et al. 2014). Additionally, a WD gene was not expressed in the wild type, and RNAi-mediated MYB10-suppressed Fv, whereas it was expressed in MYB10-overexpressing Fv fruits (Lin-Wang et al. 2014). Similarly, the putative WD gene FaAN11 was expressed more highly in FaMYB10-overexpressing flesh compared with the control, which provided a hint for exploring the interaction of MYB and WD protein (Fig. 4). Moreover, it has been suggested that a MYB TF gene (MtPAR) acts upstream of WD40-1 in Medicago truncatula (Verdier et al. 2012). All these results indicated a potential role for MYB TFs in regulation of WD expression. The roles of TFs in regulating anthocyanins accumulation The R2R3-MYB TFs play a vital role in regulation of the flavonoid pathway. It has been previously found that MYB10 was expressed at a higher level in ripe receptacles of white-fruited than in those of red-fruited Fv varieties, while the anthocyanin biosynthesis transcriptional repressor MYB1 was not included in the differentially expressed genes (Härtl et al. 2017). On the contrary, MYB10 was not differentially expressed while MYB1 was significantly down-regulated in the yellow-fruited compared with the red-fruited Fv variety according to a previous study (Zhang et al. 2015). In our results, expression of both MYB10 and MYB1 did not significantly change between red- and white-fleshed Fa cultivars. However, intriguingly, a single-repeat R3-MYB TF (MYB1R, TRINITY_DN47718_c0_g1_i1) was expressed much more highly in the W stage of white-fleshed ‘Xiaobai’, which might lead to the inhibition of anthocyanin biosynthesis in the white flesh, since MYB1R was previously known to be a negative regulator of anthocyanin accumulation. Heterologous expression of gentian MYB1R inhibited the anthocyanin accumulation in tobacco flowers (Nakatsuka et al. 2013). In addition, a putative MYB1R was up-regulated in yellow-fleshed woodland strawberry compared with its expression in red-fruited varieties (Zhang et al. 2015), suggesting that the repression of anthocyanins by high expression of MYB1R might be predominant over the activation of MYB10 on anthocyanin accumulation in strawberry. In addition, bHLH and WD TFs have been suggested as partners of R2R3-MYB TFs in activating or suppressing anthocyanin biosynthesis, which is well known as the MBW complex (Schaart et al. 2013, Li 2014, Lin-Wang et al. 2014, Xu et al. 2015). The expression level of a bHLH transcript was significantly up-regulated in white-fruited Fv genotypes compared with the red-fruited variety, while no WD was among the differentially expressed genes (Härtl et al. 2017). Similarly, we have found three bHLH transcripts differentially expressed in red-fleshed and white-fleshed strawberry. Two transcripts encoding bHLH93 were up-regulated while one transcript encoding bHLH122 was down-regulated in white-fleshed strawberry (Supplementary Table S4). However, the homolog of a previously reported MBW complex member FvbHLH33 was not found to be differentially expressed in our results or in other studies (Zhang et al. 2015, Härtl et al. 2017). In accordance with that. knock down of bHLH33 did not affect the anthocyanin pathway in Fv fruits as previously suggested (Lin-Wang et al. 2014). These results indicated that expression of bHLH33 might not be the anthocyanin biosynthesis activator responsible for the colored phenotype in the red-fruited genotypes. Notably, we found that a TTG1-like WD protein transcript FaAN11 (TRINITY_DN42135_c0_g1_i3) was significantly down-regulated in white-fleshed ‘Xiaobai’. AN11 was identified as a candidate for the fine regulation of fruit color in grapes (Costantini et al. 2015) and petunia (De Vetten et al. 1997). Its homolog in perilla was also proved to be involved in anthocyanin regulation (Sompornpailin et al. 2002). In addition to the previously known anthocyanin-related MYB, bHLH and WD TFs, recently many other TFs including WRKY have been suggested to be involved in anthocyanin regulation (Lloyd et al. 2017). The WRKY41 from Brassica napus has a similar role to WRKY41 in Arabidopsis, which could rescue the higher anthocyanin content phenotype when overexpressed in the Arabidopsis thaliana wrky41-2 mutant (Duan et al. 2018). Our results also found several differentially expressed WRKY TFs between white- and red-fleshed strawberry, consistent with the situation in potato (Liu et al. 2015c), which could help in identifying new regulators of anthocyanin biosynthesis. The participation of lncRNAs in regulating the anthocyanin pathway Recently, lncRNAs have emerged as key regulators of diverse cellular processes in mammals and plants. Deep high-throughput sequencing provides us with efficient methods for identifying lncRNAs. Thousands of lncRNAs have been identified in many species including apple (Celton et al. 2014), wheat (Xin et al. 2011) and Arabidopsis (Song et al. 2009, Liu et al. 2012, Wang et al. 2014). In our study, we identified 50,601 putative lncRNAs comprising 13.8% of the transcriptome assembles from the red- and white-fleshed Fa cultivars. This was much more than the 5,884 lncRNAs identified in Fv fruits (Kang and Liu 2015). This discrepancy might be due to the larger genome of Fa (∼720 Mb) compared with that of the Fv (∼200 Mb), similar to the circumstance in wheat as previously proposed (Sharma et al. 2017). LncRNAs can regulate gene expression through cis- and trans-action. It is reported that the vast majority (>90%) of lncRNAs may function as trans-regulators (Wang and Chang 2011, Ma et al. 2012, Lin et al. 2014). Via trans-action, lncRNAs exert their effects by regulating the expression of target genes distant from where they are transcribed (Nie et al. 2015). One such example is the lncRNA HOTAIR, which originates from the HOXC locus but could silence the HOXD locus of a different chromosome (De Lucia and Dean 2011). Co-expression analysis has been suggested as an efficient bioinformatics approach and widely used to infer trans-regulatory functions of lncRNAs (Liao et al. 2011, Zhan et al. 2016). In our results, we detected 2,070 pairs of differentially expressed lncRNAs and protein-coding transcripts involved in anthocyanin biosynthesis. Most of them showed a positive correlation, suggesting that lncRNAs could function through the trans-acting mode in strawberry anthocyanin biosynthesis. Furthermore, miRNAs can regulate gene expression at the post-transcriptional level by binding to the target sequences, resulting in cleavage, decoy or translation repression of targeted mRNA (Dalmay 2013). Over the past decades, a number of studies have uncovered the interaction among lncRNAs and miRNAs. On one hand, lncRNAs can serve as miRNA precursors for the generation of miRNAs. Thirteen lncRNAs were identified in Brassica napus as putative precursors of 96 miRNAs involved in resistance to Sclerotinia sclerotiorum infection (Joshi et al. 2016). Fourteen lncRNAs were found as precursor sequences for miRNAs in poplar under N deficiency (Chen et al. 2016). While in our study, 130 lncRNAs were identified as putative precursors of miRNAs belonging to 50 miRNA families (Supplementary Table S8). The different amount of lncRNAs as precursors of miRNAs might be due to species variation. Among the 130 lncRNAs, several lncRNAs were identified as putative precursors of miRNAs which were known as regulators of flavonoid pathways including miRNA858a, miRNA156 and miRNA396b (Gupta et al. 2017). However, no changes in expression of these lncRNAs were detected. Moreover, lncRNAs can function as the target of miRNAs, competing for binding to miRNAs (Yoon et al. 2014). Our results showed that 60 differentially expressed lncRNAs were targeted by 392 miRNAs, establishing 591 anthocyanin-responsive miRNA–lncRNA target pairs (Supplementary Table S9). All together, these results suggested that lncRNAs mostly interact with miRNAs by serving as miRNAs targets, blocking the interaction between miRNAs and their target mRNAs, and thus indirectly enhance functioning of coding transcripts by preventing negative regulation of their translation by miRNAs (Franco-Zorrilla et al. 2007). This is supported by the network and correlation analysis. For example, miRNA PC-5p-977833_4 targeted mRNA TRINITY_DN42620_c0_g1_i6 encoding CHI and also targeted TRINITY_DN48515_c0_g3_i1, which was identified as a putative lncRNA. In the white flesh of strawberry, lncRNA TRINITY_DN48515_c0_g3_i1 was down-regulated. The competition for the miRNA-binding site was reduced, resulting in the impression of TRINITY_DN42620_c0_g1_i6 by miRNA PC-5p-977833_4, and thus the lncRNA TRINITY_DN48515_c0_g3_i1 showed a positive correlation with TRINITY_DN42620_c0_g1_i6. However, a negative correlation between lncRNAs and mRNAs targeted by the same miRNA was also observed in our results. LncRNA TRINITY_DN1328_c0_g1_i1 was up-regulated in the white-fleshed strawberry. It was targeted by miRNA PC-3p-274018_13, whose target mRNA MSTRG.54516.1 encoding CHS was still down-regulated in gene expression. This might be attributed to the possibility that the protein-coding transcript MSTRG.54516.1 was also targeted by many other miRNAs and the redundancy of repression by those miRNAs. Finally, we proposed a new hypothesis elucidating the anthocyanin biosynthesis regulatory pathway and explaining the lack of anthocyanins in white-fleshed strawberry. As shown in Fig. 7, the lack of F3'H expression is the main reason for loss of cyanidin derivatives, while the down-regulation of other regulatory factors such as TFs (WD) and lncRNAs also contributed to the absence of anthocyanins in the white-fleshed strawberries. However, the truth of this matter is probably much more complex than what was proposed here; our results provided the basis for future interesting and challenging elucidation. Fig. 7 View largeDownload slide Proposed pathways for anthocyanin metabolism in different strawberry cultivars. Arrows indicate the reaction steps; the blocked steps are presented as gray and the active steps are presented as black; the main biosynthesis branches are presented as thick and the minor branches are presented as thin arrows. In the red-fleshed ‘Benihoppe’ (A), the cyanidin 3-glucoside biosynthesis branch was blocked (gray, thin arrow); pelargonidin 3-glucoside was the major class of pigment (black, thick arrow). In the flesh of FaMYB10-overexpressing ‘Xiaobai’ fruits (C), cyanidin 3-glucoside biosynthesis was recovered, accounting for about 50% of total anthocyanins. In the flesh of cv. ‘Xiaobai’ (B), both cyanidin 3-glucoside and pelargonidin 3-glucoside biosynthesis was suppressed (gray); the proposed participation of lncRNAs and miRNAs in regulating anthocyanin biosynthesis is presented as dotted lines. Purple ovals indicated putative lncRNAs; up-regulated lncRNAs were colored red and down-regulated lncRNAs were colored black. Fig. 7 View largeDownload slide Proposed pathways for anthocyanin metabolism in different strawberry cultivars. Arrows indicate the reaction steps; the blocked steps are presented as gray and the active steps are presented as black; the main biosynthesis branches are presented as thick and the minor branches are presented as thin arrows. In the red-fleshed ‘Benihoppe’ (A), the cyanidin 3-glucoside biosynthesis branch was blocked (gray, thin arrow); pelargonidin 3-glucoside was the major class of pigment (black, thick arrow). In the flesh of FaMYB10-overexpressing ‘Xiaobai’ fruits (C), cyanidin 3-glucoside biosynthesis was recovered, accounting for about 50% of total anthocyanins. In the flesh of cv. ‘Xiaobai’ (B), both cyanidin 3-glucoside and pelargonidin 3-glucoside biosynthesis was suppressed (gray); the proposed participation of lncRNAs and miRNAs in regulating anthocyanin biosynthesis is presented as dotted lines. Purple ovals indicated putative lncRNAs; up-regulated lncRNAs were colored red and down-regulated lncRNAs were colored black. In summary, we comprehensively analyzed the reasons for losing anthocyanins in white-fleshed strawberry ‘Xiaobai’, and presented the first identification of lncRNAs involved in anthocyanin regulation. Our study provided novel knowledge of anthocyanin regulation in strawberry, which may serve as important resources for future research. Materials and Methods Plant materials Strawberry (Fragaria×ananassa) ‘Benihoppe’ and its natural white-fleshed mutant ‘Xiaobai’ were grown in a greenhouse located in Shuangliu, Sichuan province, China. The growth condition was controlled at 22 ± 2°C, relative humidity 70–90% and a 14/10 h light/dark regime. Four fruit ripening stages were defined as green (G), white (W), turning (T) and full red (R) based on the days post-anthesis (DPA) and the color of the receptacle. Fruits were collected at around 18, 25, 30 and 35 DPA, respectively. The skin (outer red layer including achenes) and flesh were manually separated. Samples from three uniform fruits were subsequently ground into powder in liquid nitrogen, mixed as one biological replicate and stored at –80°C until further use. RNA extraction, library construction and RNA sequencing Total RNAs were isolated from the flesh of fruits at the W and R stage of the two cultivars using the improved CTAB (cetyltrimethylammonium bromide) method (Chen et al. 2012). RNA samples were incubated with RNase-free DNase I for 30 min to remove the genome DNA contamination. Concentration and integrity were assessed on a 1% agarose gel, also by using a NanoDrop spectrophotometer and an Agilent 2100 bioanalyzer. Finally, the samples with a concentration >400 ng μl–1, RIN (RNA integrity number) values >8 and an OD 260/280 and 260/230 ratio >1.8 were selected for library construction. rRNA was removed from the total RNA before library construction. Libraries were constructed by The Beijing Genomics Institute (BGI, Shenzhen, China). Briefly, mRNA was broken into short fragments, then first- and second-strand cDNA were synthesized. Subsequently, the cDNA was subjected to end repair and phosphorylation using T4 DNA polymerase and Klenow DNA polymerase. After that, a poly(A) tail was added at the 3′ ends of the repaired cDNA fragments and Illumina paired-end solexa adaptors were subsequently ligated to these cDNA fragments. Thereafter, the ligation products were purified on a 2% agarose gel to select a size range of templates for downstream enrichment. Next, PCR amplification was performed to enrich the purified cDNA template. Finally, the cDNA libraries were sequenced using an Illumina HiSeq™ 2000. Three independent libraries as three biological replicates were sequenced for each sample. Transcript assembly and quantification of expression Considering the potential difference between diploid and octoploid Fragaria species, we first mapped the clean reads to the Fa genome (FAN_r1.1, http://strawberry-garden.kazusa.or.jp (August 8, 2018, date last accessed)); the unmapped reads were subsequently extracted and de novo assembled using the Trinity platform (Haas et al. 2013) with the parameters of ‘min_kmer_cov = 2, normalize_reads’. Considering the incompleteness of the Fa genome and the lack of annotation information, we chose Fv proteins (https://www.ncbi.nlm.nih.gov/genome (August 8, 2018, date last accessed)) for annotation. The pooled non-redundant transcripts were further annotated by alignment against protein databases including Swiss-Prot (http://www.uniprot.org (August 8, 2018, date last accessed)) and Uniref90 (https://www.uniprot.org/help/uniref (August 8, 2018, date last accessed)) using Diamond BLASTx (Buchfink et al. 2015). Transcript expression levels were quantified and normalized by TPM values using Salmon (Patro et al. 2017). Differential expression analysis was performed using the DESeq2 R package (Love et al. 2014). Transcripts with an absolute value of log2 FC ≥1 and adjusted P-value <0.05 were defined as significantly differentially expressed. qPCR validation The expression levels of selected transcripts involved in anthocyanin biosynthesis pathways were validated by qPCR using the same RNA samples as for sequencing. Specific primers were designed using beacon 7.0 software; cDNA was synthesized using a PrimeScript RT reagent Kit with gDNA Eraser (TAKARA). SYBR Green (TAKARA) was used for detection of the PCR products on a CFX96 Real-time reaction system (Bio-Rad). The FaActin gene (Accession: LC017712) was used as the internal control for normalization of gene expression. At least two well replicates were done for each sample, and three independent RNA samples of each plant sample were used as three biological replicates to ensure reproducibility and reliability. All primers used in this study are listed in Supplementary Table S11. Measurement of anthocyanins The anthocyanins were determined using HPLC, based on a previously described method (Donno et al. 2013). Briefly, 0.2 g frozen samples were finely ground and extracted in 2 ml of extraction solution (1% HCl in methanol) for 48 h at 4°C in darkness; extraction was repeated once. Samples were centrifuged at 13,000 r.p.m. for 15 min; the supernatants were combined and filtered using a 0.45 μm Nylon filter. A 10 μl aliquot of samples was subsequently injected into the HPLC system. Compound separations were achieved on a 250 mm×4.6 mm i.d., 5 μm reversed phase Silgreen ODS C18 column (Greenherbs Science and Technology), with 95% formic acid and methanol used as mobile phases. A linear gradient (95–0%) of formic acid in methanol was used for 20 min, followed by 100% methanol for 5 min. The column temperature was kept at 25°C; the flow rate was 1 ml min–1 and chromatograms were recorded at 510 nm. Anthocyanins were quantified by comparing them with external standards. Experiments were repeated three times with three independent samples referring to the three biological replicates. Transient overexpression of FaMYB10 The full-length cDNA sequence of FaMYB10 (Accession: EU155162) was amplified from strawberry ‘Xiaobai’ and cloned into the modified pCAMBIA1301 vector with the Cauliflower mosiac virus (CaMV) 35S promoter (Supplementary Fig. S4). Agrobacterium infiltration was performed based on the previously described method (Spolaore et al. 2001). Briefly, the Agrobacterium tumefaciens strain GV3101 containing the overexpression constructs was grown at 28°C in YEB medium containing 10 mM MES (pH 5.6) and appropriate antibiotics to reach a culture OD600 of approximately 0.8, the bacteria were then concentrated and re-suspended to a final OD600 of approximately 1.0. After incubation in MMA medium, the Agrobacterium suspension was injected into the whole fruits at the W stage, when they were still attached to the plants, by a sterile 1 ml syringe. The injected fruits were harvested 6–7 d (the day they turned fully red) after injection. As a control, fruits at the same stage were injected with bacteria containing an empty vector. At least three plants with at least three fruits on each were selected for infiltration; all the fruits at the W stage on each plant were infiltrated. Identification of putative lncRNAs and correlation analysis Identification of putative lncRNAs was performed according to a pipeline (Supplementary Fig. S5). Transcripts with class_code ‘u’ (unknown intergenic transcript), ‘o’ (generic exonic overlap with a reference transcript), ‘x’ (natural antisense transcript, NAT) and ‘i’ (intronic transcript) were extracted and subjected for size and open reading frame (ORF) selection. A perl script was used to extract transcripts longer than 200 nt and shorter than 100 amino acids. As we know, a real lncRNA does not have an ORF; therefore, the longest consecutive codon chain of the lncRNAs candidates was defined as the putative ORF. The coding potential capacities of the remaining transcripts were calculated by the coding potential calculator 2 (CPC2) program (Kang et al. 2017) and PLEK software (Li et al. 2014). Only transcripts with a coding potential score of less than –0.5 and which passed the coding potential calculation were retained. Subsequently, the retained transcripts were aligned to the Fv genome annotated protein sequences (https://www.ncbi.nlm.nih.gov (August 8, 2018, date last accessed)) using Diamond software (Buchfink et al. 2015). The transcripts with identity >90%, E-value <1.0E-10 and query coverage per HSP (qcovhsp) >80% were excluded. The remaining transcripts were then subjected to the Pfam database (https://pfam.xfam.org (August 8, 2018, date last accessed)), Uniref90 and nr database (http://www.ncbi.nlm.nih.gov (August 8, 2018, date last accessed)) to exclude the known protein domain-containing transcripts by the hmmscan tool (http://hmmer.org/ (August 8, 2018, date last accessed)) using default parameters (E-value <0.001). Moreover, to eliminate any known small RNAs, these remaining transcripts were blasted against the tRNA database (http://gtrnadb.ucsc.edu/ (August 8, 2018, date last accessed)), rRNA database (https://github.com/vaulot/pr2database (August 8, 2018, date last accessed)) and miRNA database (http://www.mirbase.org (August 8, 2018, date last accessed)) (E-value <0.001). The remaining transcripts were considered as putative lncRNAs. The co-expression of lncRNAs and mRNAs was calculated by R software using the Pearson correlation algorithm; the co-expressed pairs with correlation coefficient r > 0.8 and r < –0.8 at a P-value of <0.05 were recognized as significantly correlated pairs. Construction of lncRNA–miRNA–mRNA networks Known miRNAs in strawberry were collected from the literature (Ge et al. 2013, Li et al. 2013, Xu et al. 2013) and used as query for a homology search. The possibility of lncRNAs as the precursors of miRNAs was elucidated by SUmirFind and SUmirFold perl scripts (Lucas and Budak 2012). Potential miRNAs targets were predicted by the psRNAtarget online web server (http://plantgrn.noble.org/psRNATarget/ (August 8, 2018, date last accessed)) (Dai and Zhao 2011), with default parameters. The lncRNA–miRNA–mRNA interaction networks were visualized by Cytoscape 3.4.0 software (Shannon et al. 2003). Funding This work was supported by the Sichuan Doctoral Tutor Team Supported Project. Disclosures The authors have no conflicts of interest to declare. Acknowledgements We would like to thank the Institute of Pomology & Olericulture in Sichuan Agricultural University for providing the server to analyze the transcriptome data. References Afrin S. , Gasparrini M. , Forbes-Hernandez T.Y. , Reboredo-Rodriguez P. , Mezzetti B. , Varela-López A. , et al. ( 2016 ) Promising health benefits of the strawberry: a focus on clinical studies . J. Agric. Food Chem . 64 : 4435 – 4449 . Google Scholar Crossref Search ADS PubMed Aharoni A. , De Vos C. , Wein M. , Sun Z. , Greco R. , Kroon A. , et al. ( 2001 ) The strawberry FaMYB1 transcription factor suppresses anthocyanin and flavonol accumulation in transgenic tobacco . Plant J . 28 : 319 – 332 . Google Scholar Crossref Search ADS PubMed An J.P. , Qu F.J. , Yao J.F. , Wang X.N. , You C.X. , Wang X.F. , et al. ( 2017 ) The bZIP transcription factor MdHY5 regulates anthocyanin accumulation and nitrate assimilation in apple . Hortic. Res. 4 : 17023. Google Scholar Crossref Search ADS PubMed Andersen Ø.M. , Jordheim M. ( 2010 ) Chemistry of flavonoid-based colors in plants. In Comprehensive Natural Products II. Chemistry and Biology . Edited by Mander L. , Liu H.W. pp. 547 – 614 . Elsevier Press, Oxford . Baudry A. , Heim M.A. , Dubreucq B. , Caboche M. , Weisshaar B. , Lepiniec L. ( 2004 ) TT2, TT8, and TTG1 synergistically specify the expression of BANYULS and proanthocyanidin biosynthesis in Arabidopsis thaliana . Plant J . 39 : 366 – 380 . Google Scholar Crossref Search ADS PubMed Biswas S. , Hazra S. , Chattopadhyay S. ( 2016 ) Identification of conserved miRNAs and their putative target genes in Podophyllum hexandrum (Himalayan Mayapple) . Plant Gene 6 : 82 – 89 . Google Scholar Crossref Search ADS Buchfink B. , Xie C. , Huson D.H. ( 2015 ) Fast and sensitive protein alignment using DIAMOND . Nat. Methods 12 : 59. Google Scholar Crossref Search ADS PubMed Cagirici H.B. , Alptekin B. , Budak H. ( 2017 ) RNA sequencing and co-expressed long non-coding RNA in modern and wild wheats . Sci. Rep. 7 : 10670. Google Scholar Crossref Search ADS PubMed Carbone F. , Preuss A. , De Vos R.C. , D’Amico E. , Perrotta G. , Bovy A.G. , et al. ( 2009 ) Developmental, genetic and environmental factors affect the expression of flavonoid genes, enzymes and metabolites in strawberry fruits . Plant Cell Environ . 32 : 1117 – 1131 . Google Scholar Crossref Search ADS PubMed Celton J.M. , Gaillard S. , Bruneau M. , Pelletier S. , Aubourg S. , Martin-Magniette M.L ., et al. ( 2014 ) Widespread anti-sense transcription in apple is correlated with siRNA production and indicates a large potential for transcriptional and/or post-transcriptional control . New Phytol. 203 : 287 – 299 . Google Scholar Crossref Search ADS PubMed Chekanova J.A. ( 2015 ) Long non-coding RNAs and their functions in plants . Curr. Opin. Plant Biol. 27 : 207 – 216 . Google Scholar Crossref Search ADS PubMed Chen M. , Wang C. , Bao H. , Chen H. , Wang Y. ( 2016 ) Genome-wide identification and characterization of novel lncRNAs in Populus under nitrogen deficiency . Mol. Genet. Genomics 291 : 1663 – 1680 . Google Scholar Crossref Search ADS PubMed Chen Q. , Yu H.W. , Wang X.R. , Xie X.L. , Yue X.Y. , Tang H.R. ( 2012 ) An alternative cetyltrimethylammonium bromide-based protocol for RNA isolation from blackberry (Rubus L.) . Genet. Mol. Res. 11 : 1773 – 1782 . Google Scholar Crossref Search ADS PubMed Clark S.T. , Verwoerd W.S. ( 2011 ) A systems approach to identifying correlated gene targets for the loss of colour pigmentation in plants . BMC Bioinformatics 12 : 343. Google Scholar Crossref Search ADS PubMed Costantini L. , Malacarne G. , Lorenzi S. , Troggio M. , Mattivi F. , Moser C. , et al. ( 2015 ) New candidate genes for the fine regulation of the colour of grapes . J. Exp. Bot. 66 : 4427 – 4440 . Google Scholar Crossref Search ADS PubMed Dai X. , Zhao P.X. ( 2011 ) psRNATarget: a plant small RNA target analysis server . Nucleic Acids Res. 39 : W155 – W159 . Google Scholar Crossref Search ADS PubMed Dalmay T. ( 2013 ) Mechanism of miRNA-mediated repression of mRNA translation . Essays Biochem. 54 : 29. Google Scholar Crossref Search ADS PubMed Da Silva F.L. , Escribano-Bailón M.T. , Alonso J.J.P. , Rivas-Gonzalo J.C. , Santos-Buelga C. ( 2007 ) Anthocyanin pigments in strawberry . LWT-Food Sci. Technol . 40 : 374 – 382 . Google Scholar Crossref Search ADS De Lucia F. , Dean C. ( 2011 ) Long non-coding RNAs and chromatin regulation . Curr. Opin. Plant Biol. 14 : 168 – 173 . Google Scholar Crossref Search ADS PubMed De Vetten N. , Quattrocchio F. , Mol J. , Koes R. ( 1997 ) The an11 locus controlling flower pigmentation in petunia encodes a novel WD-repeat protein conserved in yeast, plants, and animals . Genes Dev . 11 : 1422 – 1434 . Google Scholar Crossref Search ADS PubMed Donno D. , Cavanna M. , Beccaro G. , Mellano M. , Torello-Marinoni D. , Cerutti A. , et al. ( 2013 ) Currants and strawberries as bioactive compound sources: determination of antioxidant profiles with HPLC-DAD/MS . J. Appl. Bot. Food Qual . 86 : 1 – 10 . Duan S. , Wang J. , Gao C. , Jin C. , Li D. , Peng D. , et al. ( 2018 ) Functional characterization of a heterologously expressed Brassica napus WRKY41-1 transcription factor in regulating anthocyanin biosynthesis in Arabidopsis thaliana . Plant Sci . 268 : 47 – 53 . Google Scholar Crossref Search ADS PubMed Franco-Zorrilla J.M. , Valli A. , Todesco M. , Mateos I. , Puga M.I. , Rubio-Somoza I. , et al. ( 2007 ) Target mimicry provides a new mechanism for regulation of microRNA activity . Nat. Genet. 39 : 1033. Google Scholar Crossref Search ADS PubMed Ge A. , Shangguan L. , Zhang X. , Dong Q. , Han J. , Liu H. , et al. ( 2013 ) Deep sequencing discovery of novel and conserved microRNAs in strawberry (Fragaria×ananassa) . Physiol. Plant. 148 : 387 – 396 . Google Scholar Crossref Search ADS PubMed Guan L. , Dai Z. , Wu B.-H. , Wu J. , Merlin I. , Hilbert G. , et al. ( 2016 ) Anthocyanin biosynthesis is differentially regulated by light in the skin and flesh of white-fleshed and teinturier grape berries . Planta 243 : 23 – 41 . Google Scholar Crossref Search ADS PubMed Gupta O.P. , Karkute S.G. , Banerjee S. , Meena N.L. , Dahuja A. ( 2017 ) Contemporary understanding of miRNA-based regulation of secondary metabolites biosynthesis in plants . Front. Plant Sci. 8 : 374. Google Scholar PubMed Haas B.J. , Papanicolaou A. , Yassour M. , Grabherr M. , Blood P.D. , Bowden J. , et al. ( 2013 ) De novo transcript sequence reconstruction from RNA-Seq: reference generation and analysis with Trinity . Nat. Protoc. 8 : 1494. Google Scholar Crossref Search ADS PubMed Han Y. , Vimolmangkang S. , Soria-Guerra R.E. , Rosales-Mendoza S. , Zheng D. , Lygin A.V. , et al. ( 2010 ) Ectopic expression of apple F3′H genes contributes to anthocyanin accumulation in the Arabidopsis tt7 mutant grown under nitrogen stress . Plant Physiol . 153 : 806 – 820 . Google Scholar Crossref Search ADS PubMed Härtl K. , Denton A. , Franz-Oberdorf K. , Hoffmann T. , Spornraft M. , Usadel B. , et al. ( 2017 ) Early metabolic and transcriptional variations in fruit of natural white-fruited Fragaria vesca genotypes . Sci. Rep. 7 : 45113. Google Scholar Crossref Search ADS PubMed Huang H. , Hu K. , Han K.T. , Xiang Q.Y. , Dai S.L. ( 2013 ) Flower colour modification of chrysanthemum by suppression of F3'H and overexpression of the exogenous Senecio cruentus F3'5'H gene . PLoS One 8 : e74395. Google Scholar Crossref Search ADS PubMed Jiang F. , Wang J.Y. , Jia H.F. , Jia W.S. , Wang H.Q. , Xiao M. ( 2013 ) RNAi-mediated silencing of the flavanone 3-hydroxylase gene and its effect on flavonoid biosynthesis in strawberry fruit . J. Plant Growth Regul. 32 : 182 – 190 . Google Scholar Crossref Search ADS Joshi R.K. , Megha S. , Basu U. , Rahman M.H. , Kav N.N.V. ( 2016 ) Genome wide identification and functional prediction of long non-coding RNAs responsive to Sclerotinia sclerotiorum infection in Brassica napus . PLoS One 11 : e0158784 . Google Scholar Crossref Search ADS PubMed Kang C. , Liu Z. ( 2015 ) Global identification and analysis of long non-coding RNAs in diploid strawberry Fragaria vesca during flower and fruit development . BMC Genomics 16 : 815. Google Scholar Crossref Search ADS PubMed Kang Y.J. , Yang D.C. , Kong L. , Hou M. , Meng Y.Q. , Wei L. , et al. ( 2017 ) CPC2: a fast and accurate coding potential calculator based on sequence intrinsic features . Nucleic Acids Res . 45 : W12 – W16 . Google Scholar Crossref Search ADS PubMed Kosar M. , Kafkas E. , Paydas S. , Baser K.H.C. ( 2004 ) Phenolic composition of strawberry genotypes at different maturation stages . J. Agric. Food Chem. 52 : 1586 – 1589 . Google Scholar Crossref Search ADS PubMed Li A. , Zhang J. , Zhou Z. ( 2014 ) PLEK: a tool for predicting long non-coding RNAs and messenger RNAs based on an improved k-mer scheme . BMC Bioinformatics 15 : 311. Google Scholar Crossref Search ADS PubMed Li H. , Mao W. , Liu W. , Dai H. , Liu Y. , Ma Y. , et al. ( 2013 ) Deep sequencing discovery of novel and conserved microRNAs in wild type and a white-flesh mutant strawberry . Planta 238 : 695 – 713 . Google Scholar Crossref Search ADS PubMed Li S. ( 2014 ) Transcriptional control of flavonoid biosynthesis: fine-tuning of the MYB–bHLH–WD40 (MBW) complex . Plant Signal. Behav. 9 : e27522. Google Scholar Crossref Search ADS PubMed Li Z. , Rana T.M. ( 2012 ) Molecular mechanisms of RNA-triggered gene silencing machineries . Acc. Chem. Res. 45 : 1122 – 1131 . Google Scholar Crossref Search ADS PubMed Liao Q. , Liu C. , Yuan X. , Kang S. , Miao R. , Xiao H. , et al. ( 2011 ) Large-scale prediction of long non-coding RNA functions in a coding–non-coding gene co-expression network . Nucleic Acids Res . 39 : 3864 – 3878 . Google Scholar Crossref Search ADS PubMed Lin X. , Xiao M. , Luo Y. , Wang J. , Wang H. ( 2013 ) The effect of RNAi-induced silencing of FaDFR on anthocyanin metabolism in strawberry (Fragaria×ananassa) fruit . Sci. Hortic . 160 : 123 – 128 . Google Scholar Crossref Search ADS Lin X.C. , Zhu Y. , Chen W.B. , Lin L.W. , Chen D.H. , Huang J.R. , et al. ( 2014 ) Integrated analysis of long non-coding RNAs and mRNA expression profiles reveals the potential role of lncRNAs in gastric cancer pathogenesis . Int. J. Oncol . 45 : 619 – 628 . Google Scholar Crossref Search ADS PubMed Lin-Wang K. , McGhie T.K. , Wang M. , Liu Y. , Warren B. , Storey R. , et al. ( 2014 ) Engineering the anthocyanin regulatory complex of strawberry (Fragaria vesca) . Front. Plant Sci. 5 : 651. Google Scholar Crossref Search ADS PubMed Liu J. , Jung C. , Xu J. , Wang H. , Deng S. , Bernad L. , et al. ( 2012 ) Genome-wide analysis uncovers regulation of long intergenic noncoding RNAs in Arabidopsis . Plant Cell 24 : 4333 – 4345 . Google Scholar Crossref Search ADS PubMed Liu J. , Wang H. , Chua N.H. ( 2015a ) Long noncoding RNA transcriptome of plants . Plant Biotechnol. J. 13 : 319 – 328 . Google Scholar Crossref Search ADS Liu X. , Hao L. , Li D. , Zhu L. , Hu S. ( 2015b ) Long non-coding RNAs and their biological roles in plants . Genomics Proteomics Bioinformatics 13 : 137 – 147 . Google Scholar Crossref Search ADS Liu Y. , Lin-Wang K. , Deng C. , Warran B. , Wang L. , Yu B. , et al. ( 2015c ) Comparative transcriptome analysis of white and purple potato to identify genes involved in anthocyanin biosynthesis . PLoS One 10 : e0129148. Google Scholar Crossref Search ADS Lloyd A. , Brockman A. , Aguirre L. , Campbell A. , Bean A. , Cantero A. , et al. ( 2017 ) Advances in the MYB–bHLH–WD repeat (MBW) pigment regulatory model: addition of a WRKY factor and co-option of an anthocyanin MYB for betalain regulation . Plant Cell Physiol . 58 : 1431 – 1441 . Google Scholar Crossref Search ADS PubMed Lopes-da-Silva F. , de Pascual-Teresa S. , Rivas-Gonzalo J. , Santos-Buelga C. ( 2002 ) Identification of anthocyanin pigments in strawberry (cv. Camarosa) by LC using DAD and ESI-MS detection . Eur. Food Res. Technol. 214 : 248 – 253 . Google Scholar Crossref Search ADS Love M.I. , Huber W. , Anders S. ( 2014 ) Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2 . Genome Biol. 15 : 550. Google Scholar Crossref Search ADS PubMed Lucas S.J. , Budak H. ( 2012 ) Sorting the wheat from the chaff: identifying miRNAs in genomic survey sequences of Triticum aestivum chromosome 1AL . PLoS One 7 : e40859 . Google Scholar Crossref Search ADS PubMed Ma H. , Hao Y. , Dong X. , Gong Q. , Chen J. , Zhang J. , et al. ( 2012 ) Molecular mechanisms and function prediction of long noncoding RNA . ScientificWorldJournal 2012 : 541786 Google Scholar PubMed Mazza G. , Miniati E. ( 1993 ) Anthocyanins in Fruits, Vegetables, and Grains . CRC Press , Boca Raton, FL . Medina-Puche L. , Cumplido-Laso G. , Amil-Ruiz F. , Hoffmann T. , Ring L. , Rodríguez-Franco A. , et al. ( 2014 ) MYB10 plays a major role in the regulation of flavonoid/phenylpropanoid metabolism during ripening of Fragaria×ananassa fruits . J. Exp. Bot. 65 : 401 – 417 . Google Scholar Crossref Search ADS PubMed Mochizuki Y. , Iwasaki Y. , Fuke M. , Ogiwara I. ( 2014 ) Analysis of a high-yielding strawberry (Fragaria×ananassa Duch.) cultivar ‘Benihoppe’ with focus on root dry matter and activity . J. Jpn. Soc. Hort. Sci. 83 : 142 – 148 . Google Scholar Crossref Search ADS Nakatsuka T. , Yamada E. , Saito M. , Fujita K. , Nishihara M. ( 2013 ) Heterologous expression of gentian MYB1R transcription factors suppresses anthocyanin pigmentation in tobacco flowers . Plant Cell Rep. 32 : 1925 – 1937 . Google Scholar Crossref Search ADS PubMed Nie M. , Deng Z.-L. , Liu J. , Wang D.Z. ( 2015 ) Noncoding RNAs, emerging regulators of skeletal muscle development and diseases . BioMed Res. Int . 2015 : 676575. Google Scholar PubMed Nithin C. , Thomas A. , Basak J. , Bahadur R.P. ( 2017 ) Genome-wide identification of miRNAs and lncRNAs in Cajanus cajan . BMC Genomics 18 : 878 . Google Scholar Crossref Search ADS PubMed Patro R. , Duggal G. , Love M.I. , Irizarry R.A. , Kingsford C. ( 2017 ) Salmon provides fast and bias-aware quantification of transcript expression . Nat. Methods 14 : 417 – 419 . Google Scholar Crossref Search ADS PubMed Pillet J. , Folta K.M. ( 2015 ) Pigments in strawberry. In Pigments in Fruits and Vegetables . Edited by Chen C. pp. 205 – 216 . Springer Press , New York . Reddy A.M. , Reddy V.S. , Scheffler B.E. , Wienand U. , Reddy A.R. ( 2007 ) Novel transgenic rice overexpressing anthocyanidin synthase accumulates a mixture of flavonoids leading to an increased antioxidant potential . Metab. Eng . 9 : 95 – 111 . Google Scholar Crossref Search ADS PubMed Salvatierra A. , Pimentel P. , Moya-Leon M.A. , Caligari P.D. , Herrera R. ( 2010 ) Comparison of transcriptional profiles of flavonoid genes and anthocyanin contents during fruit development of two botanical forms of Fragaria chiloensis ssp. chiloensis . Phytochemistry 71 : 1839 – 1847 . Google Scholar Crossref Search ADS PubMed Schaart J.G. , Dubos C. , Romero De La Fuente I. , Houwelingen A.M. , Vos R.C. , Jonker H.H. , et al. ( 2013 ) Identification and characterization of MYB–bHLH–WD40 regulatory complexes controlling proanthocyanidin biosynthesis in strawberry (Fragaria×ananassa) fruits . New Phytol. 197 : 454 – 467 . Google Scholar Crossref Search ADS PubMed Shafiq S. , Li J. , Sun Q. ( 2016 ) Functions of plants long non-coding RNAs . Biochim. Biophys. Acta. 1859 : 155. Google Scholar Crossref Search ADS PubMed Shannon P. , Markiel A. , Ozier O. , Baliga N.S. , Wang J.T. , Ramage D. , et al. ( 2003 ) Cytoscape: a software environment for integrated models of biomolecular interaction networks . Genome Res . 13 : 2498 – 2504 . Google Scholar Crossref Search ADS PubMed Sharma S. , Taneja M. , Tyagi S. , Singh K. , Upadhyay S.K. ( 2017 ) Survey of high throughput RNA-seq data reveals potential roles for lncRNAs during development and stress response in bread wheat . Front. Plant Sci . 8 : 1019 . Google Scholar Crossref Search ADS PubMed Shin D.H. , Choi M.-G. , Kang C.S. , Park C.-S. , Choi S.B. , Park Y.I. ( 2016 ) Overexpressing the wheat dihydroflavonol 4-reductase gene TaDFR increases anthocyanin accumulation in an Arabidopsis dfr mutant . Genes Genom. 38 : 333 – 340 . Google Scholar Crossref Search ADS Sompornpailin K. , Makita Y. , Yamazaki M. , Saito K. ( 2002 ) A WD-repeat-containing putative regulatory protein in anthocyanin biosynthesis in Perilla frutescens . Plant Mol. Biol . 50 : 485 – 495 . Google Scholar Crossref Search ADS PubMed Sondheimer E. , Karash C.B. ( 1956 ) The major anthocyanin pigments of the wild strawberry (Fragaria vesca) . Nature 178 : 648 – 649 . Google Scholar Crossref Search ADS Song D. , Yang Y. , Yu B. , Zheng B. , Deng Z. , Lu B.L. , et al. ( 2009 ) Computational prediction of novel non-coding RNAs in Arabidopsis thaliana. BMC Bioinformatics 10 : S36 . Google Scholar Crossref Search ADS PubMed Spolaore S. , Trainotti L. , Casadoro G. ( 2001 ) A simple protocol for transient gene expression in ripe fleshy fruit mediated by agrobacterium . J. Exp. Bot. 52 : 845 – 850 . Google Scholar Crossref Search ADS PubMed Verdier J. , Zhao J. , Torres-Jerez I. , Ge S. , Liu C. , He X. , et al. ( 2012 ) MtPAR MYB transcription factor acts as an on switch for proanthocyanidin biosynthesis in Medicago truncatula . Proc. Natl. Acad. Sci. USA 109 : 1766 – 1771 . Google Scholar Crossref Search ADS Wang J. , Meng X. , Dobrovolskaya O.B. , Orlov Y.L. , Chen M. ( 2017 ) Non-coding RNAs and their roles in stress response in plants . Genomics Proteomics Bioinformatics 15 : 301 – 312 . Google Scholar Crossref Search ADS PubMed Wang K.C. , Chang H.Y. ( 2011 ) Molecular mechanisms of long noncoding RNAs . Mol. Cell. 43 : 904 – 914 . Google Scholar Crossref Search ADS PubMed Wang S.Y. , Zheng W. , Galletta G.J. ( 2002 ) Cultural system affects fruit quality and antioxidant capacity in strawberries . J. Agric. Food Chem. 50 : 6534 – 6542 . Google Scholar Crossref Search ADS PubMed Wang Y. , Wang X. , Deng W. , Fan X. , Liu T.-T. , He G. , et al. ( 2014 ) Genomic features and regulatory roles of intermediate-sized non-coding RNAs in Arabidopsis . Mol. Plant. 7 : 514 – 527 . Google Scholar Crossref Search ADS PubMed Wu X. , Gong Q. , Ni X. , Zhou Y. , Gao Z. ( 2017 ) UFGT: the key enzyme associated with the petals variegation in Japanese apricot . Front. Plant Sci. 8 : 108. Google Scholar PubMed Wu X. , Prior R.L. ( 2005 ) Systematic identification and characterization of anthocyanins by HPLC-ESI-MS/MS in common foods in the United States: fruits and berries . J. Agric. Food Chem. 53 : 2589 – 2599 . Google Scholar Crossref Search ADS PubMed Xie S. , Song C. , Wang X. , Liu M. , Zhang Z. , Xi Z. ( 2015 ) Tissue-specific expression analysis of anthocyanin biosynthetic genes in white- and red-fleshed grape cultivars . Molecules 20 : 22767 – 22780 . Google Scholar Crossref Search ADS PubMed Xin M. , Wang Y. , Yao Y. , Song N. , Hu Z. , Qin D. , et al. ( 2011 ) Identification and characterization of wheat long non-protein coding RNAs responsive to powdery mildew infection and heat stress by using microarray analysis and SBS sequencing . BMC Plant Biol. 11 : 61. Google Scholar Crossref Search ADS PubMed Xu W. , Dubos C. , Lepiniec L. ( 2015 ) Transcriptional control of flavonoid biosynthesis by MYB–bHLH–WDR complexes . Trends Plant Sci. 20 : 176 – 185 . Google Scholar Crossref Search ADS PubMed Xu X. , Yin L. , Ying Q. , Song H. , Xue D. , Lai T. , et al. ( 2013 ) High-throughput sequencing and degradome analysis identify miRNAs and their targets involved in fruit senescence of Fragaria ananassa . PLoS One 8 : e70959. Google Scholar Crossref Search ADS PubMed Yoon J.H. , Abdelmohsen K. , Gorospe M. ( 2014 ) Functional interactions among microRNAs and long noncoding RNAs . Semin. Cell Dev. Biol. 34 : 9 – 14 . Google Scholar Crossref Search ADS PubMed Zhan S. , Dong Y. , Zhao W. , Guo J. , Zhong T. , Wang L. , et al. ( 2016 ) Genome-wide identification and characterization of long non-coding RNAs in developmental skeletal muscle of fetal goat . BMC Genomics 17 : 666. Google Scholar Crossref Search ADS PubMed Zhang Y. , Li W. , Dou Y. , Zhang J. , Jiang G. , Miao L. , et al. ( 2015 ) Transcript quantification by RNA-Seq reveals differentially expressed genes in the red and yellow fruits of Fragaria vesca . PLoS One 10 : e0144356. Google Scholar Crossref Search ADS PubMed Zhu B. , Yang Y. , Li R. , Fu D. , Wen L. , Luo Y. , et al. ( 2015 ) RNA sequencing and functional analysis implicate the regulatory role of long non-coding RNAs in tomato fruit ripening . J. Exp. Bot. 66 : 4483 – 4495 . Google Scholar Crossref Search ADS PubMed Zufall R.A. , Rausher M.D. ( 2004 ) Genetic changes associated with floral adaptation restrict future evolutionary potential . Nature 428 : 847 – 850 . Google Scholar Crossref Search ADS PubMed Abbreviations Abbreviations ANR anthocyanidin reductase ANS anthocyanidin synthase bHLH basic helix–loop–helix bZIP basic leucine zipper CHI chalcone isomerase CHS chalcone synthase CPC2 coding potential calculator 2 DFR dihydroflavonol 4-reductase DPA days post-anthesis ERF ethylene-responsive factor Fa Fragaria×ananassa FC fold change F3H flavanone 3-hydroxylase F3'H flavonoid 3'-hydroxylase FLS flavonol synthase Fv Fragaria vesca G stage green stage LAR leucoanthocyanidin reductase lncRNA long non-coding RNA MBW MYB–bHLH–WD40 miRNA microRNA ncRNA non-coding RNA ORF open reading frame qPCR real-time quantitative PCR R stage full red stage RNAi RNA interference RNAseq RNA sequencing T stage turning stage TF transcription factor TPM transcripts per kilobase million UFGT anthocyanidin 3-O-glucosylthransferase W stage white stage © The Author(s) 2018. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Plant and Cell Physiology Oxford University Press

Comparative Transcriptome Profiling Analysis of Red- and White-Fleshed Strawberry (Fragaria×ananassa) Provides New Insight into the Regulation of the Anthocyanin Pathway

Loading next page...
 
/lp/ou_press/comparative-transcriptome-profiling-analysis-of-red-and-white-fleshed-4dkdYOiSMM
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: journals.permissions@oup.com
ISSN
0032-0781
eISSN
1471-9053
D.O.I.
10.1093/pcp/pcy098
Publisher site
See Article on Publisher Site

Abstract

Abstract Anthocyanins are water-soluble pigments in plants. They confer both economic and healthy profits for humans. To gain a deeper insight into the regulation of anthocyanin biosynthesis in octoploid strawberry (Fragaria×ananassa; Fa), a widely consumed economically important fruit, we performed comparative transcriptomic analysis of red- and white-fleshed strawberry cultivars in two ripening stages. In total, 365,455 non-redundant transcripts were assembled from the RNA sequencing (RNAseq) data. Of this collection, 377 were annotated as putative anthocyanin-related transcripts. Differential expression analysis revealed that 57 anthocyanin biosynthesis transcripts were down-regulated, and 89 transcription factors (TFs) were either down- or up-regulated under anthocyanin deficiency. Additionally, amongst the 50,601 putative long non-coding RNAs (lncRNAs) identified here, 68 lncRNAs were differentially expressed and co-expressed with differentially expressed anthocyanin-related mRNAs; 2,070 co-expressing lncRNA–mRNA pairs were generated. Expression profile analysis revealed that it was the limited expression of FaF3'H (flavonoid 3'-hydroxylase) that blocked the cyanidin 3-glucoside accumulation in the two investigated strawberry cultivars. This was further supported by a transient overexpression experiment with FaMYB10. The down-regulated lncRNAs might participate in anthocyanin regulation by acting as targets for microRNAs (miRNAs). The level of competitive intensity in miRNA and lncRNA for the same mRNA targets was probably lower in the white-fleshed strawberries, which can release the repression effect of the mRNAs in red-fleshed strawberry as a result. This study for the first time presents lncRNAs related to anthocyanins in strawberries, provides new insights into the anthocyanin regulatory network and also lays the foundation for identifying new anthocyanin regulators in strawberry. Introduction Strawberry (Fragaria×ananassa; Fa) is widely consumed not only for its enriched bioactive compounds but also for its attractive fruit color. These key quality traits are attributed to the compounds such as anthocyanins which are recognized as one of the most important antioxidants, thereby contributing to the healthful attributes (Afrin et al. 2016). Different types and content of anthocyanins bring us different colored strawberry cultivars ranging from orange to extremely dark red (Pillet and Folta 2015). Moreover, white, yellow, peach and pink-blushed strawberries also exist, suggesting distinct anthocyanin metabolism among cultivars. It is necessary to investigate the underlying metabolism in strawberries in order to improve fruit quality. To this end, the accessions with different colors, especially those deficient in anthocyanin accumulation, provide us with good opportunities. Anthocyanins are water-soluble pigments, belonging to the flavonoid class. They are glycosides or acylglycosides of polyhydroxyl, derivatives of 2-phenylbenzopyrylium or flavylium salts. Nearly 700 different anthocyanins have been identified so far (Andersen and Jordheim 2010). In strawberry, numerous anthocyanins have been identified, including 3-rutinosides of pelargonidin and cyanidin (Da Silva et al. 2007), pelargonidin 3-(malonyl) glucoside and pelargonidin 3-(6-acetyl)-glucoside (Wu and Prior 2005), cyanidin 3-(succinoyl) glucoside and pelargonidin 3-(succinoyl) glucoside (Wang et al. 2002). The pelargonidin 3-glucoside and cyanidin 3-glucoside, conferring bright and dark red color, respectively, have been recognized as two major anthocyanins (Sondheimer and Karash 1956, Mazza and Miniati 1993), while pelargonidin 3-glucoside occurs in a much higher amount than cyanidin 3-glucoside. It accounts for >70% of total anthocyanins (Wang et al. 2002, Kosar et al. 2004, Donno et al. 2013). Anthocyanins are derived from the branched flavonoid biosynthetic pathway. It is proven that the accumulation of anthocyanins is regulated by the expression levels of biosynthetic genes, including chalcone synthase (CHS), chalcone isomerase (CHI), flavanone 3-hydroxylase (F3H), flavonoid 3'-hydroxylase (F3'H), dihydroflavonol 4-reductase (DFR) and anthocyanidin synthase (ANS). (Salvatierra et al. 2010). If any step of these enzymatic catabolites is up-regulated or blocked/perturbed this can lead to variations in the final products, and then in the visual color. For instance, overexpression of F3'H from apple in Arabidopsis and tobacco enhanced the accumulation of anthocyanins and resulted in red seedlings and flowers, respectively (Han et al. 2010), while silencing of DFR (Lin et al. 2013) or F3H (Jiang et al. 2013) in strawberry has resulted in complete loss of red color in fruits. Besides these enzymatic reaction-related structural genes, transcription factors (TFs) are proved to be essential regulators in anthocyanin biosynthesis. Among them, MYB, basic helix–loop–helix (bHLH) and WD40-repeat (WD) proteins are the most extensively investigated types. They act independently or co-operate with each other as a ternary MYB–bHLH–WD40 (MBW) complex (Baudry et al. 2004, Xu et al. 2015) to regulate anthocyanin accumulation. In strawberry, several members of the MBW complex have been identified and characterized as regulators controlling proanthocyanidin biosynthesis (Schaart et al. 2013). What is more, knock-down of FvMYB10 resulted in undetectable concentrations of anthocyanins in woodland strawberry (Fragaria vesca; Fv), while the FvMYB10 overexpression lines had significantly elevated the anthocyanin levels (Lin-Wang et al. 2014), suggesting the pivotal role of MYB10 on regulating anthocyanin accumulation. This is also supported by the study describing a lack of anthocyanin production in strawberry when FaMYB10 was transiently silenced (Medina-Puche et al. 2014). On the other hand, knock-down of FvbHLH33 in woodland strawberry showed no effect on anthocyanin accumulation (Lin-Wang et al. 2014). In addition, repressors of anthocyanins were also identified, such as FaMYB1 (Aharoni et al. 2001), FaMYB5 and FabHLH3Δ (Schaart et al. 2013). More recently, many other TF families have been demonstrated to be involved in anthocyanin modulation, such as basic leucine zipper (bZIP) (An et al. 2017) and WRKY (Duan et al. 2018), suggesting the existence of complex regulatory networks. However, information about the regulation of anthocyanins by such TFs in strawberry is limited; more novel TFs need to be identified and characterized. Long non-coding RNAs (lncRNAs) are a class of functional RNAs with a length longer than 200 nt, lacking protein-coding capacity. They were originally thought to be transcriptional ‘noise’, due to their low expression and high sequence conservation compared with protein-coding mRNAs (Chekanova 2015, Liu et al. 2015a, Liu et al. 2015b, Shafiq et al. 2016). However, more and more evidence has been found to indicate that lncRNAs play critical regulatory roles in diverse biological processes in plants, including stress response (Wang et al. 2017), flower and fruit development (Zhu et al. 2015) and ripening (Kang and Liu 2015). In the past decades, numerous studies have been carried out to explore the functions of lncRNAs. It has been suggested that lncRNAs can regulate gene expression in both the cis- and trans-acting mode, exerting their functions on neighboring genes on the same allele, or distant alleles far away from where they were transcribed (Li and Rana 2012). Moreover, lncRNAs can interact with micro RNAs (miRNAs) by serving either as precursors to generate miRNAs, or as targets of miRNAs to compete for the miRNA binding with mRNA targets (Wang et al. 2017). As another type of non-coding RNAs (ncRNAs), some miRNAs have been reported to regulate the flavonoid and anthocyanin biosynthesis pathways (Gupta et al. 2017). For example, miR829.1 and miR1873 have been computationally identified in Podophyllum heandrum to target mRNAs coding for CHS and DFR, respectively (Biswas et al. 2016). Taken together, these results suggested that lncRNAs might participate in regulating anthocyanin biosynthesis. Although lncRNAs have been identified from several species, such as wheat (Cagirici et al. 2017) and pigeon pea (Nithin et al. 2017), information about lncRNAs and their expression profile response to anthocyanin deficiency in strawberry is still lacking. Through deep sequencing, the overall picture of regulation of the anthocyanin biosynthesis pathway can be investigated. To unravel the complex network of regulation of anthocyanin biosynthesis in strawberry, we generated transcriptomic profiling of anthocyanin-related genes in red-fleshed strawberry and its natural anthocyanin-deficient mutant (white-fleshed strawberries). Comparative transcriptomic analysis was conducted to assess why anthocyanins are deficient in the flesh of white-fleshed strawberries. Moreover, we carried out transcriptome-scale identification of lncRNAs in strawberry. The expression profiles of lncRNAs in response to anthocyanin deficiency were also analyzed. In addition, the potential regulatory network of lncRNAs–miRNAs–mRNAs was constructed, which gave us important clues regarding the roles of lncRNAs in regulating anthocyanin biosynthesis, providing new insights into anthocyanin regulation. Results Anthocyanin accumulation during strawberry fruit ripening We performed HPLC analysis of the major anthocyanins (cyanidin 3-glucoside and pelargonidin 3-glucoside) in the two strawberry cultivars (Fig. 1A). Strawberry ‘Benihoppe’ has red skin and red flesh; it is the progeny resulting from a cross between cv. ‘Akihime’ and cv. ‘Sachinoka’ (Mochizuki et al. 2014). ‘Xiaobai’ differs from ‘Benihoppe’ during tissue culture selection; it has red skin but white flesh. Since the two cultivars have similarly colored skin and different colored flesh, we examined the anthocyanins in skin (outer red layer including achenes) and flesh separately. As expected, anthocyanins were detected in the skin and flesh of ‘Benihoppe’ with a higher level in the fruit skin than in the flesh (Table 1). Anthocyanins were also detected in the skin of ‘Xiaobai’. On the other hand, no anthocyanins were detected in the white flesh of ‘Xiaobai’ during the whole fruit ripening process, from green (G) stage to the full red (R) stage (Table 1), which implied that anthocyanin biosynthesis has been blocked in the flesh of the cultivar as we can see from the phenotype (Fig. 1A). Further, the types of accumulated anthocyanins were analyzed. Interestingly, even in the red flesh of cv. ‘Benihoppe’, no cyanidin 3-glucoside accumulation was detected (Fig. 1B). Pelargonidin 3-glucoside was detected as the major anthocyanin type (Fig. 1B). However, in the skin, both ‘Benihoppe’ and ‘Xiaobai’ produced cyanidin 3-glucoside (Table 1, Fig. 1B), indicating different anthocyanin metabolism in strawberry skin and flesh. Table 1 Anthocyanin content (μg g-1 FW)a during strawberry fruit ripening ‘Benihoppe’ skin ‘Benihoppe’ flesh ‘Xiaobai’ skin ‘Xiaobai’ flesh Stages Cy Pg Cy Pg Cy Pg Cy Pg Green ND ND ND ND ND ND ND ND White ND ND ND ND ND ND ND ND Turning 30.4 ± 2.0 75.7 ± 18.5 ND 53.4 ± 10.2 26.2 ± 0.6 45.2 ± 5.2 ND ND Half red 44.0 ± 6.4 187.4 ± 21.1 ND 89.0 ± 6.1 32.2 ± 3.1 91.0 ± 4.1 ND ND Full red 43.8 ± 4.0 325.2 ± 37.6 ND 123.0 ± 15.1 37.6 ± 2.4 125.1 ± 6.0 ND ND ‘Benihoppe’ skin ‘Benihoppe’ flesh ‘Xiaobai’ skin ‘Xiaobai’ flesh Stages Cy Pg Cy Pg Cy Pg Cy Pg Green ND ND ND ND ND ND ND ND White ND ND ND ND ND ND ND ND Turning 30.4 ± 2.0 75.7 ± 18.5 ND 53.4 ± 10.2 26.2 ± 0.6 45.2 ± 5.2 ND ND Half red 44.0 ± 6.4 187.4 ± 21.1 ND 89.0 ± 6.1 32.2 ± 3.1 91.0 ± 4.1 ND ND Full red 43.8 ± 4.0 325.2 ± 37.6 ND 123.0 ± 15.1 37.6 ± 2.4 125.1 ± 6.0 ND ND a Anthocyanin content is represented as the average value ± SD of three biological replicates. Cy, cyanidin 3-glucoside; Pg, pelargonidin 3-glucoside; ND, not detected. Table 1 Anthocyanin content (μg g-1 FW)a during strawberry fruit ripening ‘Benihoppe’ skin ‘Benihoppe’ flesh ‘Xiaobai’ skin ‘Xiaobai’ flesh Stages Cy Pg Cy Pg Cy Pg Cy Pg Green ND ND ND ND ND ND ND ND White ND ND ND ND ND ND ND ND Turning 30.4 ± 2.0 75.7 ± 18.5 ND 53.4 ± 10.2 26.2 ± 0.6 45.2 ± 5.2 ND ND Half red 44.0 ± 6.4 187.4 ± 21.1 ND 89.0 ± 6.1 32.2 ± 3.1 91.0 ± 4.1 ND ND Full red 43.8 ± 4.0 325.2 ± 37.6 ND 123.0 ± 15.1 37.6 ± 2.4 125.1 ± 6.0 ND ND ‘Benihoppe’ skin ‘Benihoppe’ flesh ‘Xiaobai’ skin ‘Xiaobai’ flesh Stages Cy Pg Cy Pg Cy Pg Cy Pg Green ND ND ND ND ND ND ND ND White ND ND ND ND ND ND ND ND Turning 30.4 ± 2.0 75.7 ± 18.5 ND 53.4 ± 10.2 26.2 ± 0.6 45.2 ± 5.2 ND ND Half red 44.0 ± 6.4 187.4 ± 21.1 ND 89.0 ± 6.1 32.2 ± 3.1 91.0 ± 4.1 ND ND Full red 43.8 ± 4.0 325.2 ± 37.6 ND 123.0 ± 15.1 37.6 ± 2.4 125.1 ± 6.0 ND ND a Anthocyanin content is represented as the average value ± SD of three biological replicates. Cy, cyanidin 3-glucoside; Pg, pelargonidin 3-glucoside; ND, not detected. Fig. 1 View largeDownload slide Phenotypes and HPLC analysis of anthocyanins in red- and white-fleshed strawberries. (A) Phenotypes of red-fleshed strawberry ‘Benihoppe’ (left) and white-fleshed strawberry ‘Xiaobai’ (right). Black arrows indicate the consistent position where the flesh was collected for further experiments. (B) HPLC analysis of anthocyanins in the skin and flesh of red- and white-fleshed strawberries. Fig. 1 View largeDownload slide Phenotypes and HPLC analysis of anthocyanins in red- and white-fleshed strawberries. (A) Phenotypes of red-fleshed strawberry ‘Benihoppe’ (left) and white-fleshed strawberry ‘Xiaobai’ (right). Black arrows indicate the consistent position where the flesh was collected for further experiments. (B) HPLC analysis of anthocyanins in the skin and flesh of red- and white-fleshed strawberries. RNAseq and assembly The flesh of fruits at the white (W) stage and R stage of each cultivar with three biological replicates was used to build 12 libraries for high-throughput RNA sequencing (RNAseq). After removing the low-quality reads, around 100,000,000 clean reads of each sample were obtained. Overall, 55% of these reads could be mapped to the Fa strawberry reference genome (Supplementary Table S1). The unmapped reads with the corresponding mates were extracted and then de novo assembled. Finally, we obtained 365,455 non-redundant transcripts in total, with an average length of 643 bp, which were used for further analysis. Expression profiling of anthocyanin biosynthesis genes To investigate in which step anthocyanin synthesis was blocked in the white flesh of ‘Xiaobai’, we identified the candidate transcripts for anthocyanin biosynthesis by searching the standard gene names and mapping all transcripts to reference pathways in the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (Supplementary Fig. S1). In total, 377 transcripts involved in three pigment-related pathways, i.e. flavonoid biosynthesis, anthocyanin biosynthesis, and flavone and flavonol biosynthesis pathways, were identified (Table 2). The homologs in Fv are listed in Supplementary Table S2. Secondly, the abundances of these anthocyanin biosynthesis transcripts were estimated (Supplementary Table S2) and compared in white flesh and red flesh of strawberries. The transcripts with per kilobase million (TPM) value <2 across all samples were filtered out. As expected, all of the involved transcripts showed significantly down-regulated patterns in the white flesh of ‘Xiaobai’ compared with the red flesh of ‘Benihoppe’ strawberries. The only exception was observed in FaF3'H and two FaFLS transcripts, which showed no change or up-regulation in expression, respectively (Supplementary Table S3). The normalized abundances of differentially expressed transcripts were shown by the heatmap in Fig. 2. Moreover, a large proportion of these transcripts were inhibited at both the W stage and R stage in the white flesh of ‘Xiaobai’, while some were only down-regulated at either the W stage or R stage (Table 2; Supplementary Table S3). For instance, the expression of two transcripts encoding CHS (MSTRG.75989.1) and F3H (MSTRG.78844.1) was inhibited at both the W and R stage in the white flesh compared with the red flesh. Another CHS transcript (MSTRG.54516.2) was only down-regulated in the W stage, and one CHI transcript (TRINITY_DN43706_c1_g1_i1) was only down-regulated in the R stage. Table 2 Candidate genes related to anthocyanins in strawberry Pathways Genes Enzymes KO ids No. alla No. upb No. downc W R W R Anthocyanin biosynthesis CHS Chalcone synthase K00660 38 0 0 11 9 CHI Chalcone isomerase K01859 24 0 0 11 6 F3H Flavanone 3-hydroxylase K00475 20 0 0 13 13 F3'H Flavonoid 3'-hydroxylase K05280 6 0 0 0 0 DFR Dihydroflavonol 4-reductase K13082 61 0 0 9 4 ANS Anthocyanidin synthase K05277 17 0 0 7 4 UFGT Anthocyanidin 3-O-glucosylthransferase K12930 51 0 0 8 7 Anthocyanin modification GT1 Anthocyanidin 5,3-O-glucosyltransferase K12938 9 0 0 0 1 3GGT anthocyanidin 3-O-glucoside 2''-O-glucosyltransferase K12939 6 0 0 0 0 AA5GT Cyanidin 3-O-glucoside 5-O-glucosyltransferase (acyl-glucose) K17194 2 0 0 0 0 FAOMT Flavonoid 3',5'-methyltransferase K13272 3 0 0 0 0 5MaT1 Anthocyanin 5-O-glucoside-6''-O-melonytransferase K12934 1 0 0 0 0 UF3GT Flavonoid 3-O-glucosyltransferase K17193 86 0 0 1 0 Flavone and flavonol FLS Flavonol synthase K05278 16 0 2 0 4 Flavonone biosynthesis ANR Anthocyanidin reductase K08695 30 0 0 0 0 LAR Leucoanthocyanidin reductase K13081 7 0 0 0 0 Pathways Genes Enzymes KO ids No. alla No. upb No. downc W R W R Anthocyanin biosynthesis CHS Chalcone synthase K00660 38 0 0 11 9 CHI Chalcone isomerase K01859 24 0 0 11 6 F3H Flavanone 3-hydroxylase K00475 20 0 0 13 13 F3'H Flavonoid 3'-hydroxylase K05280 6 0 0 0 0 DFR Dihydroflavonol 4-reductase K13082 61 0 0 9 4 ANS Anthocyanidin synthase K05277 17 0 0 7 4 UFGT Anthocyanidin 3-O-glucosylthransferase K12930 51 0 0 8 7 Anthocyanin modification GT1 Anthocyanidin 5,3-O-glucosyltransferase K12938 9 0 0 0 1 3GGT anthocyanidin 3-O-glucoside 2''-O-glucosyltransferase K12939 6 0 0 0 0 AA5GT Cyanidin 3-O-glucoside 5-O-glucosyltransferase (acyl-glucose) K17194 2 0 0 0 0 FAOMT Flavonoid 3',5'-methyltransferase K13272 3 0 0 0 0 5MaT1 Anthocyanin 5-O-glucoside-6''-O-melonytransferase K12934 1 0 0 0 0 UF3GT Flavonoid 3-O-glucosyltransferase K17193 86 0 0 1 0 Flavone and flavonol FLS Flavonol synthase K05278 16 0 2 0 4 Flavonone biosynthesis ANR Anthocyanidin reductase K08695 30 0 0 0 0 LAR Leucoanthocyanidin reductase K13081 7 0 0 0 0 a No. all: the total number of transcripts that were analyzed. b No. up: the number of transcripts that were significantly up-regulated (|log2 FCe| |log2 FC| >1, FDR <0.05) in white-fleshed ‘Xiaobai’ compared with red-fleshed ‘Benihoppe’. c No. down: the number of transcripts that were significantly down-regulated (|log2 FC| >1, FDR <0.05) in white-fleshed ‘Xiaobai’ compared with red-fleshed ‘Benihoppe’. W, white stage; R, full red stage. Table 2 Candidate genes related to anthocyanins in strawberry Pathways Genes Enzymes KO ids No. alla No. upb No. downc W R W R Anthocyanin biosynthesis CHS Chalcone synthase K00660 38 0 0 11 9 CHI Chalcone isomerase K01859 24 0 0 11 6 F3H Flavanone 3-hydroxylase K00475 20 0 0 13 13 F3'H Flavonoid 3'-hydroxylase K05280 6 0 0 0 0 DFR Dihydroflavonol 4-reductase K13082 61 0 0 9 4 ANS Anthocyanidin synthase K05277 17 0 0 7 4 UFGT Anthocyanidin 3-O-glucosylthransferase K12930 51 0 0 8 7 Anthocyanin modification GT1 Anthocyanidin 5,3-O-glucosyltransferase K12938 9 0 0 0 1 3GGT anthocyanidin 3-O-glucoside 2''-O-glucosyltransferase K12939 6 0 0 0 0 AA5GT Cyanidin 3-O-glucoside 5-O-glucosyltransferase (acyl-glucose) K17194 2 0 0 0 0 FAOMT Flavonoid 3',5'-methyltransferase K13272 3 0 0 0 0 5MaT1 Anthocyanin 5-O-glucoside-6''-O-melonytransferase K12934 1 0 0 0 0 UF3GT Flavonoid 3-O-glucosyltransferase K17193 86 0 0 1 0 Flavone and flavonol FLS Flavonol synthase K05278 16 0 2 0 4 Flavonone biosynthesis ANR Anthocyanidin reductase K08695 30 0 0 0 0 LAR Leucoanthocyanidin reductase K13081 7 0 0 0 0 Pathways Genes Enzymes KO ids No. alla No. upb No. downc W R W R Anthocyanin biosynthesis CHS Chalcone synthase K00660 38 0 0 11 9 CHI Chalcone isomerase K01859 24 0 0 11 6 F3H Flavanone 3-hydroxylase K00475 20 0 0 13 13 F3'H Flavonoid 3'-hydroxylase K05280 6 0 0 0 0 DFR Dihydroflavonol 4-reductase K13082 61 0 0 9 4 ANS Anthocyanidin synthase K05277 17 0 0 7 4 UFGT Anthocyanidin 3-O-glucosylthransferase K12930 51 0 0 8 7 Anthocyanin modification GT1 Anthocyanidin 5,3-O-glucosyltransferase K12938 9 0 0 0 1 3GGT anthocyanidin 3-O-glucoside 2''-O-glucosyltransferase K12939 6 0 0 0 0 AA5GT Cyanidin 3-O-glucoside 5-O-glucosyltransferase (acyl-glucose) K17194 2 0 0 0 0 FAOMT Flavonoid 3',5'-methyltransferase K13272 3 0 0 0 0 5MaT1 Anthocyanin 5-O-glucoside-6''-O-melonytransferase K12934 1 0 0 0 0 UF3GT Flavonoid 3-O-glucosyltransferase K17193 86 0 0 1 0 Flavone and flavonol FLS Flavonol synthase K05278 16 0 2 0 4 Flavonone biosynthesis ANR Anthocyanidin reductase K08695 30 0 0 0 0 LAR Leucoanthocyanidin reductase K13081 7 0 0 0 0 a No. all: the total number of transcripts that were analyzed. b No. up: the number of transcripts that were significantly up-regulated (|log2 FCe| |log2 FC| >1, FDR <0.05) in white-fleshed ‘Xiaobai’ compared with red-fleshed ‘Benihoppe’. c No. down: the number of transcripts that were significantly down-regulated (|log2 FC| >1, FDR <0.05) in white-fleshed ‘Xiaobai’ compared with red-fleshed ‘Benihoppe’. W, white stage; R, full red stage. Fig. 2 View largeDownload slide Schematic of the anthocyanin biosynthesis pathway in strawberry. The expression patterns of transcripts were represented by the log2 TPM value and shown as a heatmap at the side of each step. The four cells from left to right represent the white stage of the red-fleshed cv. (W-R), the red stage of the red-fleshed cv. (R-R), the white stage of the white-fleshed cv. (W-W) and the red stage of the white-fleshed cv. (R-W), respectively. PAL, phenylammonia-lyase; C4H, cinnamate-4-hydroxylase; 4CL, 4-coumaroyl-CoA synthase; CHS, chalcone synthase; CHI, chalcone isomerase; F3H, flavanone 3-hydroxylase; F3′H, flavonoid 3′-hydroxylase; ANS, anthocyanidin synthase; FLS, flavonol synthase; DFR, dihydroflavonol 4-reductase; UFGT, anthocyanidin 3-O-glucosyltransferase; LAR, leucoanthocyanidin reductase; ANR, anthocyanidin reductase. Fig. 2 View largeDownload slide Schematic of the anthocyanin biosynthesis pathway in strawberry. The expression patterns of transcripts were represented by the log2 TPM value and shown as a heatmap at the side of each step. The four cells from left to right represent the white stage of the red-fleshed cv. (W-R), the red stage of the red-fleshed cv. (R-R), the white stage of the white-fleshed cv. (W-W) and the red stage of the white-fleshed cv. (R-W), respectively. PAL, phenylammonia-lyase; C4H, cinnamate-4-hydroxylase; 4CL, 4-coumaroyl-CoA synthase; CHS, chalcone synthase; CHI, chalcone isomerase; F3H, flavanone 3-hydroxylase; F3′H, flavonoid 3′-hydroxylase; ANS, anthocyanidin synthase; FLS, flavonol synthase; DFR, dihydroflavonol 4-reductase; UFGT, anthocyanidin 3-O-glucosyltransferase; LAR, leucoanthocyanidin reductase; ANR, anthocyanidin reductase. White flesh was recovered to red by overexpression of FaMYB10 To discover in depth the molecular reasons for losing color in the white flesh of strawberry ‘Xiaobai’, we successfully turned the white flesh back into red flesh by transient overexpression of one FaMYB10 TF gene (Fig. 3A). Cyanidin 3-glucoside was detected at a higher level in the entire FaMYB10-overexpressing ‘Xiaobai’ fruits (including both skin and flesh) than in the control (fruits infiltrated with bacteria containing empty vector) (Fig. 3B;Supplementary Fig. S2). This was different from what we have described in the previous section, i.e. that no cyanidin 3-glucoside was detected either in the red flesh of ‘Benihoppe’ or in the white flesh of ‘Xiaobai’ fruits (Fig. 1), indicating that FaMYB10 recovered the cyanidin 3-glucoside biosynthesis in strawberry. This proposal was also supported by overexpression of FaMYB10 in ‘Benihoppe’, as we also detected a higher concentration of cyanidin 3-glucoside in the flesh of FaMYB10-overexpressing‘Benihoppe’ fruits (Fig. 3B;Supplementary Fig. S2). To elucidate this matter, the expression levels of some key genes for anthocyanin biosynthesis in the red flesh of FaMYB10-overexpressing ‘Xiaobai’ fruits were estimated by real-time quantitative PCR (qPCR). As shown in Fig. 4, anthocyanin biosynthesis transcripts including FaDFR, FaANS and anthocyanidin 3-O-glucosylthransferase (FaUFGT) accumulated at a much higher level in the flesh of FaMYB10-overexpressing ‘Xiaobai’ fruits compared with the control (Fig. 4). Specially, the FaF3'H gene was expressed in the flesh of FaMYB10-overexpressing fruits, while it was rarely expressed in the normal red flesh of ‘Benihoppe’ and the white flesh of ‘Xiaobai’ fruits (Fig. 2; Supplementary Table S2). These findings revealed the key reason why cyanidin 3-glucoside accumulated in the flesh of FaMYB10-overexpressing fruits. Also, expression levels of TFs related to anthocyanin biosynthesis including WD and bHLH33 were estimated in the FaMYB10-overexpressing flesh. An increase in WD expression but no change in bHLH33 expression was observed, indicating a potential regulatory role for FaMYB10 in WD expression. Fig. 3 View largeDownload slide Phenotypes and anthocyanin content of FaMYB10-overexpressing strawberry. (A) The phenotypes of FaMYB10-overexpressing ‘Benihoppe’ (up) and ‘Xiaobai’ (down). White arrows indicate the position where the flesh color was changed after infiltration. (B) Anthocyanin content in FaMYB10-overexpressing fruits. 35s, control fruits infiltrated with Agrobacterium containing an empty vector; 35s:FaMYB10, experimental fruits infiltrated with Agrobacterium containing the recombined FaMYB10 overexpression plasmid. Fig. 3 View largeDownload slide Phenotypes and anthocyanin content of FaMYB10-overexpressing strawberry. (A) The phenotypes of FaMYB10-overexpressing ‘Benihoppe’ (up) and ‘Xiaobai’ (down). White arrows indicate the position where the flesh color was changed after infiltration. (B) Anthocyanin content in FaMYB10-overexpressing fruits. 35s, control fruits infiltrated with Agrobacterium containing an empty vector; 35s:FaMYB10, experimental fruits infiltrated with Agrobacterium containing the recombined FaMYB10 overexpression plasmid. Fig. 4 View largeDownload slide The expression of anthocyanin-related genes detected by qPCR in strawberry flesh. Bars represent the average relative transcript expression; error bars represent the SD of three biological replicates. R-W, the red stage of the white-fleshed cv. ‘Xiaobai’; R-R, the red stage of the red-fleshed cv. ‘Benihoppe’; 35s, control fruits infiltrated with Agrobacterium containing an empty vector; 35s:FaMYB10, experimental fruits infiltrated with Agrobacterium containing recombined FaMYB10 overexpression plasmid. Fig. 4 View largeDownload slide The expression of anthocyanin-related genes detected by qPCR in strawberry flesh. Bars represent the average relative transcript expression; error bars represent the SD of three biological replicates. R-W, the red stage of the white-fleshed cv. ‘Xiaobai’; R-R, the red stage of the red-fleshed cv. ‘Benihoppe’; 35s, control fruits infiltrated with Agrobacterium containing an empty vector; 35s:FaMYB10, experimental fruits infiltrated with Agrobacterium containing recombined FaMYB10 overexpression plasmid. qPCR validation of RNAseq To verify the gene expression data obtained by RNAseq analysis, eight transcripts involved in the anthocyanin biosynthesis pathway were selected for qPCR using the same RNA samples as used for sequencing (Supplementary Fig. S3). The results showed that despite the quantitative differences in expression levels, the expression patterns detected by qPCR experiments of most transcripts were consistent with the expression patterns investigated by RNAseq analysis. A linear correlation between the qPCR and the RNAseq data was observed (R2 = 0.73) (Fig. 5), indicating the reliability of the data. Fig. 5 View largeDownload slide Correlation analysis of qPCR relative expression and RNAseq fold change. Fig. 5 View largeDownload slide Correlation analysis of qPCR relative expression and RNAseq fold change. Identification of anthocyanin-related TFs As the structural genes involved in anthocyanin biosynthesis are largely regulated at the transcriptional level, we identified 89 TFs either significantly up- or down-regulated in W and R stages of white flesh compared with the red flesh strawberries (Supplementary Table S4). This collection included one MYB, three bHLH and three WD TFs. Intriguingly, a transcript (TRINITY_DN47718_c0_g1_i1) annotated as a single-repeat R3 type MYB (MYB1R) was much more highly expressed in the W stage of white-fleshed ‘Xiaobai’, with a log2 fold change (FC) (white/red) value of 8.6. In addition, a TTG1-like WD transcript, FaAN11 (TRINITY_DN42135_c0_g1_i3), exhibited significant down-regulation in the white flesh of ‘Xiaobai’ compared with the red flesh of ‘Benihoppe’, with a log2 FC (white/red) value of –5.1 in the R stage. In additon, many other TF types including ethylene-responsive factors (ERFs), WRKYs and NAC domain-containing proteins were also included in the differentially expressed TF sets (Supplementary Table S4), indicating the possibility of novel TFs participating in anthocyanin regulation. Characteristics of putative lncRNAs and their expression under anthocyanin deficiency Putative lncRNAs were identified based on our transcriptome data. First, among the 365,445 non-redundant transcripts, 299,935 transcripts with class_code ‘u’, ‘o’, ‘x’ and ‘i’ were extracted, and subsequently 241,307 transcripts passed the defined length criteria. A total of 54,244 transcripts were identified as non-coding transcripts after filtering by the coding potential calculator 2 (CPC2) server, and through the PLEK software. Potential protein domain-containing transcripts among these transcripts were removed by blasting against the Fv protein, Pfam and nr databases. The remaining 51,446 transcripts were kept for further analysis. To eliminate any known small RNAs, these transcripts were blasted against the tRNA database, rRNA database and miRNA database. Finally, 13.8% (50,601 transcripts) of the transcriptome assemblies were identified as putative lncRNAs (Supplementary Table S5). The length of lncRNAs ranged from 201 to 2,585 nt. The average length was 381 nt, and the average GC content was around 40%, with the highest GC content detected in shorter lncRNAs. A total of 41,570 transcripts with a TPM value >2 in at least three samples were identified as actively expressed putative lncRNAs (Supplementary Table S6). Differential expression analysis showed that when compared with red flesh from ‘Benihoppe’, 19 putative lncRNAs were significantly down-regulated, 12 were up-regulated at the W stage, and 49 and 20 lncRNAs were down-regulated or up-regulated at the R stage, respectively (Supplementary Table S7), in the white flesh of ‘Xiaobai’. Six and four transcripts were down- or up-regulated at both the R and W stage, respectively. The largest log2 FC (white/red) value was shown by TRINITY_DN1328_c1_g1_i1 (–10.1 at the R stage) and TRINITY_DN23526_c0_g1_i2 (13.29 at the W stage). Co-expression of lncRNAs and anthocyanin-related coding transcripts or TFs We carried out Pearson relative correlation analysis to investigate the trans-co-expression of differentially expressed lncRNAs, coding transcripts and TFs involved in the anthocyanin biosynthesis pathway. As a result, 64 lncRNAs interacted with 74 coding transcripts and TFs, establishing 2,070 lncRNA–mRNA pairs that were differentially co-expressed between white flesh and red flesh at a significant level (Supplementary Table S8). Most down-regulated lncRNAs showed a positive correlation, while up-regulated lncRNAs negatively correlated with mRNAs, except for FLS transcripts. Further, we observed that a particular lncRNA could be co-expressed with multiple coding transcripts and TFs, and multiple lncRNAs have been co-expressed with one particular coding transcript or TF. TRINITY_DN50364_c2_g1_i3 and TRINITY_DN47222_c0_g2_i7 appeared to be the most co-expressed lncRNA, being down-regulated and up-regulated, respectively. Construction of lncRNA–miRNA–mRNA networks Literature records have revealed that lncRNA functions are associated with miRNAs by acting as either their precursors or their targets. To investigate the details of this, 1,584 published miRNAs from strawberry were collected for network analysis. As a result, 130 lncRNAs were identified as putative precursors of 80 miRNAs (Supplementary Table S9). Interestingly, only one (TRINITY_DN50364_c2_g1_i3) of the precursor lncRNAs exhibited a down-regulated expression pattern in white-fleshed ‘Xiaobai’. It was detected as the precursor of miRNA PC-5p-142026_25, which also targeted one lncRNA transcript, TRINITY_DN49666_c1_g2_i21, showing no change in expression between the white- and red-fleshed cultivars. Moreover, lots of lncRNAs were identified as putative precursors of miRNAs which were previously known as regulators of flavonoid pathways, such as miRNA858a, miRNA156 and miRNA396b. However, no expression changes were detected in these lncRNAs. In order to provide more insights into lncRNA–miRNA association, the lncRNAs targets of miRNAs were predicted. The results showed that 60 differentially expressed lncRNAs were targeted by 392 miRNAs, establishing 591 anthocyanin-responsive miRNA–lncRNA target pairs (Supplementary Table S10). Further, the differentially expressed lncRNAs and their miRNA–mRNA network related to anthocyanins was visualized (Fig. 6). Among the interaction networks pairs, several miRNAs targeted one lncRNA and one mRNA, such as PC-5p-977833_4, fan-miR478, fan-miR2630, PC-5p-469001_8 and PC-5p-439332_9. In contrast, fan-miR2630, PC-3p-228442_16, PC-3p-217435_17, PC-5p-56438_63 and S4-m0037 targeted one lncRNA and several mRNAs. LncRNAs and mRNAs targeted by the same miRNAs were down-regulated in the white flesh of ‘Xiaobai’. Specifically, the up-regulated lncRNA TRINITY_DN44094_c0_g1_i15, TRINITY_DN1328_c0_g1_i1 and TRINITY_DN50080_c2_g3_i1 were targeted by miRNA PC-5p-2988_1087 and fan-miR869, respectively. Both miRNAs have one or more down-regulated mRNA targets. The down-regulated lncRNA TRINITY_DN43354_c0_g4_i1 was targeted by the miRNA PC-5p-150653_24, whose anthocyanin-related mRNA targets have been up-regulated. All these results indicated the complex interaction of lncRNAs, miRNAs and mRNAs. Fig. 6 View largeDownload slide Network between differentially expressed lncRNAs and anthocyanin-related protein-coding transcripts regulated by miRNAs. miRNA nodes were represented as triangles; lncRNAs and coding transcripts were represented as rectangles and circles, respectively. Transcripts that were up-regulated in white-fleshed ‘Xiaobai’ were colored red, and down-regulated transcripts were colored black. Fig. 6 View largeDownload slide Network between differentially expressed lncRNAs and anthocyanin-related protein-coding transcripts regulated by miRNAs. miRNA nodes were represented as triangles; lncRNAs and coding transcripts were represented as rectangles and circles, respectively. Transcripts that were up-regulated in white-fleshed ‘Xiaobai’ were colored red, and down-regulated transcripts were colored black. Discussion The role of FaF3'H in strawberry cyanidin 3-glucoside biosynthesis Anthocyanins are glycosides or acylglycosides of polyhydroxyl. Developmental programming plays a vital role in anthocyanin biosynthesis in strawberry. It was even thought to be the predominant factor over genotypes and environmental factors (Carbone et al. 2009). In support of this, our results showed that anthocyanin accumulation started from the T stage (the stage at which fruits start becoming red) and gradually increased with fruit ripening. The full red fruits had the highest level of anthocyanins (Table 1). In addition, anthocyanin accumulation exhibits tissue-specific characteristics. As previously suggested (Guan et al. 2016), the regulation of anthocyanin biosynthesis by light in the skin was different from that in the flesh of white-fleshed and teinturier grape berries. This might be caused by the differential expression of anthocyanin biosynthesis genes in different tissues. As an example, the anthocyanin biosynthesis genes (F3'5'H and ANS) were found to be differentially expressed in the skin and flesh of red-fleshed grapes (Xie et al. 2015). Additionally, in our results, cyanidin 3-glucoside was detected in the skin of both red-fleshed and white-fleshed strawberries, whereas in the flesh, no cyanidin 3-glucoside was detected. This finding was different from the previous reports describing that cyanidin 3-glucoside was detected at a considerable level in strawberry (Sondheimer and Karash 1956, Mazza and Miniati 1993, Lopes-da-Silva et al. 2002, Da Silva et al. 2007). This difference could be attributed to the differences in varieties. However, we detected cyanidin 3-glucoside accumulation in the flesh of FaMYB10-overexpressing fruits, suggesting that FaMYB10 has recovered the biosynthesis of cyanidin 3-glucoside. To elucidate in depth, we focused on the cyanidin biosynthesis branch. The two genes involved in this branch, ANS and UFGT, were expressed at a much higher level in the FaMYB10-overexpressing flesh compared with the control (Fig. 4). This difference might contribute to the cyanidin synthesis in FaMYB10-overexpressing fruits. According to this scenario, the red-fleshed ‘Benihoppe’ fruits with higher expression of these genes than the white-fleshed ‘Xiaobai’ should form cyanidin derivatives. However, even though these genes were more highly expressed in red-fleshed ‘Benihoppe’ than in white-fleshed ‘Xiaobai’ with the highest FC (red/white) of 7 (UFGT), the cyanidin derivatives are still absent in the flesh of red ‘Benihoppe’ fruits. Yet, it is not a satisfactory explanation for the lack of cyanidin derivatives in the white flesh of ‘Xiaobai’ strawberries. Therefore, F3'H, a necessary gene for formatting cyanidin, was considered. As generally known, F3'H catalyzes the first step in this branch. It was proven that a mutation in the coding region of the F3'H gene resulted in its inability to produce cyanidin-derived anthocyanins in morning glory (Zufall and Rausher 2004). Suppression of F3'H by RNA interference (RNAi) in chrysanthemum successfully shifted the predominantly cyanidin-derived pigments to non-pigment transgenic lines (Huang et al. 2013). On the other hand, overexpression of F3'H from apple in Arabidopsis resulted in more anthocyanin accumulation in mutants than in the wild type (Han et al. 2010). These results indicated the vital role of F3'H in anthocyanin accumulation. Our results showed that FaF3'H was rarely expressed in the red flesh of ‘Benihoppe’ and white flesh of ‘Xiaobai’, both of which had a lack of cyanidin derivatives. Hoever, it was expressed at a much higher level in FaMYB10-overexpressing flesh (Fig. 4), accompanied by a high level of cyanidin 3-glucoside accumulation. It is reasonable to conclude that the low expression of FaF3'H blocked the cyanidin 3-glucoside synthesis in the two studieded strawberry cultivars. The roles of anthocyanin biosynthesis genes in regulating anthocyanin accumulation Anthocyanin accumulation corresponds to the expression of genes encoding enzymes involved in the biosynthesis pathway. It has been previously suggested that major genes involved in anthocyanin biosynthesis were inhibited in a yellow strawberry variety at the turning stage (Zhang et al. 2015). Our study here showed that almost all key genes in anthocyanin biosynthesis were down-regulated from the W stage, indicating that the anthocyanin biosynthesis transcripts are perturbed at early developmental stages. This finding is consistent with conclusions that gene expression perturbation of flavonoid pathway genes in a white stawberry variety occurs in the early green developmental stage (Härtl et al. 2017). In our results, the expression levels of early biosynthetic genes such as CHS and F3H have been severely inhibited by 2- to 4-fold. The ‘late’ genes in anthocyanin biosynthesis, including DFR, ANS and UFGT, were also suppressed, with the largest log2 FC (red/white) around 7 in white-fleshed ‘Xioabai’. CHS catalyzes the first reaction step in forming the primary precursor for anthocyanin biosynthesis. If the expression of CHS was strongly inhibited, the anthocyanin concentration in strawberry fruits was significantly reduced, which finally led to the loss of fruit pigment accompanied by an increasing lignin content (Clark and Verwoerd 2011). F3H converts flavanone into dihydroflavanol. The absolute depletion of red color in F3H-silenced strawberry fruits revealed the critical role of F3H in blocking the anthocyanin biosynthesis (Jiang et al. 2013). DFR catalyzes the conversion of dihydroflavonol into colorless leucoanthocyanidins. Suppression of DFR in strawberry resulted in reduction of fruit red pigment (Lin et al. 2013), while overexpression of DFR could rescue the dfr mutant phenotype in Arabidopsis (Shin et al. 2016). ANS affects anthocyanin accumulation directly by converting the colorless leucoanthocyanidins to colored anthocyanidins (Reddy et al. 2007). However, we tried transient overexpression of one FaANS gene (accession: AY695817) in white-fleshed ‘Xiaobai’ strawberries; no pigment accumulated in the FaANS-overexpressing fruits (data not shown), suggesting that the lower expression of this structural gene might not be responsible for the loss-of-color ‘Xiaobai’ phenotype. However, we could not exclude its contribution to modulating anthocyanin content. In addition, UFGT catalyzes the key step for anthocyanin stability and water solubility in plants. The expression of UFGT was detected at a much higher level in red-fleshed than in white-fleshed cultivars, which is similar to the results observed in grape skin (Wu et al. 2017), indicating the role of UFGT in regulating anthocyanins. The regulation by FaMYB10 of anthocyanin pathways It has been reported that ANS is not regulated by transient silencing of FaMYB10 in Fa (Medina-Puche et al. 2014). Similar results have been found in RNAi FaMYB10 knocked-down Fv fruits (Lin-Wang et al. 2014). In contrast, we found that the expression levels of ANS were increased in FaMYB10-overexpressing flesh (Fig. 4). This finding is similar to the previous results showing that in the FaMYB10 overexpression lines, the expression of ANS was significantly up-regulated (Lin-Wang et al. 2014). This indicated the complex regulatory role of FaMYB10 in ANS expression. On the other hand, the expression of transcripts encoding TF FabHLH33 exhibited no change in FaMYB10-overexpressing fruits, indicating that FaMYB10 could not regulate the expression of FabHLH33 in Fa strawberry. This was supported by the fact that the expression of FvbHLH33 did not change either in FvMYB10-overexpressing or in FvMYB10-silenced Fv fruits (Lin-Wang et al. 2014). Additionally, a WD gene was not expressed in the wild type, and RNAi-mediated MYB10-suppressed Fv, whereas it was expressed in MYB10-overexpressing Fv fruits (Lin-Wang et al. 2014). Similarly, the putative WD gene FaAN11 was expressed more highly in FaMYB10-overexpressing flesh compared with the control, which provided a hint for exploring the interaction of MYB and WD protein (Fig. 4). Moreover, it has been suggested that a MYB TF gene (MtPAR) acts upstream of WD40-1 in Medicago truncatula (Verdier et al. 2012). All these results indicated a potential role for MYB TFs in regulation of WD expression. The roles of TFs in regulating anthocyanins accumulation The R2R3-MYB TFs play a vital role in regulation of the flavonoid pathway. It has been previously found that MYB10 was expressed at a higher level in ripe receptacles of white-fruited than in those of red-fruited Fv varieties, while the anthocyanin biosynthesis transcriptional repressor MYB1 was not included in the differentially expressed genes (Härtl et al. 2017). On the contrary, MYB10 was not differentially expressed while MYB1 was significantly down-regulated in the yellow-fruited compared with the red-fruited Fv variety according to a previous study (Zhang et al. 2015). In our results, expression of both MYB10 and MYB1 did not significantly change between red- and white-fleshed Fa cultivars. However, intriguingly, a single-repeat R3-MYB TF (MYB1R, TRINITY_DN47718_c0_g1_i1) was expressed much more highly in the W stage of white-fleshed ‘Xiaobai’, which might lead to the inhibition of anthocyanin biosynthesis in the white flesh, since MYB1R was previously known to be a negative regulator of anthocyanin accumulation. Heterologous expression of gentian MYB1R inhibited the anthocyanin accumulation in tobacco flowers (Nakatsuka et al. 2013). In addition, a putative MYB1R was up-regulated in yellow-fleshed woodland strawberry compared with its expression in red-fruited varieties (Zhang et al. 2015), suggesting that the repression of anthocyanins by high expression of MYB1R might be predominant over the activation of MYB10 on anthocyanin accumulation in strawberry. In addition, bHLH and WD TFs have been suggested as partners of R2R3-MYB TFs in activating or suppressing anthocyanin biosynthesis, which is well known as the MBW complex (Schaart et al. 2013, Li 2014, Lin-Wang et al. 2014, Xu et al. 2015). The expression level of a bHLH transcript was significantly up-regulated in white-fruited Fv genotypes compared with the red-fruited variety, while no WD was among the differentially expressed genes (Härtl et al. 2017). Similarly, we have found three bHLH transcripts differentially expressed in red-fleshed and white-fleshed strawberry. Two transcripts encoding bHLH93 were up-regulated while one transcript encoding bHLH122 was down-regulated in white-fleshed strawberry (Supplementary Table S4). However, the homolog of a previously reported MBW complex member FvbHLH33 was not found to be differentially expressed in our results or in other studies (Zhang et al. 2015, Härtl et al. 2017). In accordance with that. knock down of bHLH33 did not affect the anthocyanin pathway in Fv fruits as previously suggested (Lin-Wang et al. 2014). These results indicated that expression of bHLH33 might not be the anthocyanin biosynthesis activator responsible for the colored phenotype in the red-fruited genotypes. Notably, we found that a TTG1-like WD protein transcript FaAN11 (TRINITY_DN42135_c0_g1_i3) was significantly down-regulated in white-fleshed ‘Xiaobai’. AN11 was identified as a candidate for the fine regulation of fruit color in grapes (Costantini et al. 2015) and petunia (De Vetten et al. 1997). Its homolog in perilla was also proved to be involved in anthocyanin regulation (Sompornpailin et al. 2002). In addition to the previously known anthocyanin-related MYB, bHLH and WD TFs, recently many other TFs including WRKY have been suggested to be involved in anthocyanin regulation (Lloyd et al. 2017). The WRKY41 from Brassica napus has a similar role to WRKY41 in Arabidopsis, which could rescue the higher anthocyanin content phenotype when overexpressed in the Arabidopsis thaliana wrky41-2 mutant (Duan et al. 2018). Our results also found several differentially expressed WRKY TFs between white- and red-fleshed strawberry, consistent with the situation in potato (Liu et al. 2015c), which could help in identifying new regulators of anthocyanin biosynthesis. The participation of lncRNAs in regulating the anthocyanin pathway Recently, lncRNAs have emerged as key regulators of diverse cellular processes in mammals and plants. Deep high-throughput sequencing provides us with efficient methods for identifying lncRNAs. Thousands of lncRNAs have been identified in many species including apple (Celton et al. 2014), wheat (Xin et al. 2011) and Arabidopsis (Song et al. 2009, Liu et al. 2012, Wang et al. 2014). In our study, we identified 50,601 putative lncRNAs comprising 13.8% of the transcriptome assembles from the red- and white-fleshed Fa cultivars. This was much more than the 5,884 lncRNAs identified in Fv fruits (Kang and Liu 2015). This discrepancy might be due to the larger genome of Fa (∼720 Mb) compared with that of the Fv (∼200 Mb), similar to the circumstance in wheat as previously proposed (Sharma et al. 2017). LncRNAs can regulate gene expression through cis- and trans-action. It is reported that the vast majority (>90%) of lncRNAs may function as trans-regulators (Wang and Chang 2011, Ma et al. 2012, Lin et al. 2014). Via trans-action, lncRNAs exert their effects by regulating the expression of target genes distant from where they are transcribed (Nie et al. 2015). One such example is the lncRNA HOTAIR, which originates from the HOXC locus but could silence the HOXD locus of a different chromosome (De Lucia and Dean 2011). Co-expression analysis has been suggested as an efficient bioinformatics approach and widely used to infer trans-regulatory functions of lncRNAs (Liao et al. 2011, Zhan et al. 2016). In our results, we detected 2,070 pairs of differentially expressed lncRNAs and protein-coding transcripts involved in anthocyanin biosynthesis. Most of them showed a positive correlation, suggesting that lncRNAs could function through the trans-acting mode in strawberry anthocyanin biosynthesis. Furthermore, miRNAs can regulate gene expression at the post-transcriptional level by binding to the target sequences, resulting in cleavage, decoy or translation repression of targeted mRNA (Dalmay 2013). Over the past decades, a number of studies have uncovered the interaction among lncRNAs and miRNAs. On one hand, lncRNAs can serve as miRNA precursors for the generation of miRNAs. Thirteen lncRNAs were identified in Brassica napus as putative precursors of 96 miRNAs involved in resistance to Sclerotinia sclerotiorum infection (Joshi et al. 2016). Fourteen lncRNAs were found as precursor sequences for miRNAs in poplar under N deficiency (Chen et al. 2016). While in our study, 130 lncRNAs were identified as putative precursors of miRNAs belonging to 50 miRNA families (Supplementary Table S8). The different amount of lncRNAs as precursors of miRNAs might be due to species variation. Among the 130 lncRNAs, several lncRNAs were identified as putative precursors of miRNAs which were known as regulators of flavonoid pathways including miRNA858a, miRNA156 and miRNA396b (Gupta et al. 2017). However, no changes in expression of these lncRNAs were detected. Moreover, lncRNAs can function as the target of miRNAs, competing for binding to miRNAs (Yoon et al. 2014). Our results showed that 60 differentially expressed lncRNAs were targeted by 392 miRNAs, establishing 591 anthocyanin-responsive miRNA–lncRNA target pairs (Supplementary Table S9). All together, these results suggested that lncRNAs mostly interact with miRNAs by serving as miRNAs targets, blocking the interaction between miRNAs and their target mRNAs, and thus indirectly enhance functioning of coding transcripts by preventing negative regulation of their translation by miRNAs (Franco-Zorrilla et al. 2007). This is supported by the network and correlation analysis. For example, miRNA PC-5p-977833_4 targeted mRNA TRINITY_DN42620_c0_g1_i6 encoding CHI and also targeted TRINITY_DN48515_c0_g3_i1, which was identified as a putative lncRNA. In the white flesh of strawberry, lncRNA TRINITY_DN48515_c0_g3_i1 was down-regulated. The competition for the miRNA-binding site was reduced, resulting in the impression of TRINITY_DN42620_c0_g1_i6 by miRNA PC-5p-977833_4, and thus the lncRNA TRINITY_DN48515_c0_g3_i1 showed a positive correlation with TRINITY_DN42620_c0_g1_i6. However, a negative correlation between lncRNAs and mRNAs targeted by the same miRNA was also observed in our results. LncRNA TRINITY_DN1328_c0_g1_i1 was up-regulated in the white-fleshed strawberry. It was targeted by miRNA PC-3p-274018_13, whose target mRNA MSTRG.54516.1 encoding CHS was still down-regulated in gene expression. This might be attributed to the possibility that the protein-coding transcript MSTRG.54516.1 was also targeted by many other miRNAs and the redundancy of repression by those miRNAs. Finally, we proposed a new hypothesis elucidating the anthocyanin biosynthesis regulatory pathway and explaining the lack of anthocyanins in white-fleshed strawberry. As shown in Fig. 7, the lack of F3'H expression is the main reason for loss of cyanidin derivatives, while the down-regulation of other regulatory factors such as TFs (WD) and lncRNAs also contributed to the absence of anthocyanins in the white-fleshed strawberries. However, the truth of this matter is probably much more complex than what was proposed here; our results provided the basis for future interesting and challenging elucidation. Fig. 7 View largeDownload slide Proposed pathways for anthocyanin metabolism in different strawberry cultivars. Arrows indicate the reaction steps; the blocked steps are presented as gray and the active steps are presented as black; the main biosynthesis branches are presented as thick and the minor branches are presented as thin arrows. In the red-fleshed ‘Benihoppe’ (A), the cyanidin 3-glucoside biosynthesis branch was blocked (gray, thin arrow); pelargonidin 3-glucoside was the major class of pigment (black, thick arrow). In the flesh of FaMYB10-overexpressing ‘Xiaobai’ fruits (C), cyanidin 3-glucoside biosynthesis was recovered, accounting for about 50% of total anthocyanins. In the flesh of cv. ‘Xiaobai’ (B), both cyanidin 3-glucoside and pelargonidin 3-glucoside biosynthesis was suppressed (gray); the proposed participation of lncRNAs and miRNAs in regulating anthocyanin biosynthesis is presented as dotted lines. Purple ovals indicated putative lncRNAs; up-regulated lncRNAs were colored red and down-regulated lncRNAs were colored black. Fig. 7 View largeDownload slide Proposed pathways for anthocyanin metabolism in different strawberry cultivars. Arrows indicate the reaction steps; the blocked steps are presented as gray and the active steps are presented as black; the main biosynthesis branches are presented as thick and the minor branches are presented as thin arrows. In the red-fleshed ‘Benihoppe’ (A), the cyanidin 3-glucoside biosynthesis branch was blocked (gray, thin arrow); pelargonidin 3-glucoside was the major class of pigment (black, thick arrow). In the flesh of FaMYB10-overexpressing ‘Xiaobai’ fruits (C), cyanidin 3-glucoside biosynthesis was recovered, accounting for about 50% of total anthocyanins. In the flesh of cv. ‘Xiaobai’ (B), both cyanidin 3-glucoside and pelargonidin 3-glucoside biosynthesis was suppressed (gray); the proposed participation of lncRNAs and miRNAs in regulating anthocyanin biosynthesis is presented as dotted lines. Purple ovals indicated putative lncRNAs; up-regulated lncRNAs were colored red and down-regulated lncRNAs were colored black. In summary, we comprehensively analyzed the reasons for losing anthocyanins in white-fleshed strawberry ‘Xiaobai’, and presented the first identification of lncRNAs involved in anthocyanin regulation. Our study provided novel knowledge of anthocyanin regulation in strawberry, which may serve as important resources for future research. Materials and Methods Plant materials Strawberry (Fragaria×ananassa) ‘Benihoppe’ and its natural white-fleshed mutant ‘Xiaobai’ were grown in a greenhouse located in Shuangliu, Sichuan province, China. The growth condition was controlled at 22 ± 2°C, relative humidity 70–90% and a 14/10 h light/dark regime. Four fruit ripening stages were defined as green (G), white (W), turning (T) and full red (R) based on the days post-anthesis (DPA) and the color of the receptacle. Fruits were collected at around 18, 25, 30 and 35 DPA, respectively. The skin (outer red layer including achenes) and flesh were manually separated. Samples from three uniform fruits were subsequently ground into powder in liquid nitrogen, mixed as one biological replicate and stored at –80°C until further use. RNA extraction, library construction and RNA sequencing Total RNAs were isolated from the flesh of fruits at the W and R stage of the two cultivars using the improved CTAB (cetyltrimethylammonium bromide) method (Chen et al. 2012). RNA samples were incubated with RNase-free DNase I for 30 min to remove the genome DNA contamination. Concentration and integrity were assessed on a 1% agarose gel, also by using a NanoDrop spectrophotometer and an Agilent 2100 bioanalyzer. Finally, the samples with a concentration >400 ng μl–1, RIN (RNA integrity number) values >8 and an OD 260/280 and 260/230 ratio >1.8 were selected for library construction. rRNA was removed from the total RNA before library construction. Libraries were constructed by The Beijing Genomics Institute (BGI, Shenzhen, China). Briefly, mRNA was broken into short fragments, then first- and second-strand cDNA were synthesized. Subsequently, the cDNA was subjected to end repair and phosphorylation using T4 DNA polymerase and Klenow DNA polymerase. After that, a poly(A) tail was added at the 3′ ends of the repaired cDNA fragments and Illumina paired-end solexa adaptors were subsequently ligated to these cDNA fragments. Thereafter, the ligation products were purified on a 2% agarose gel to select a size range of templates for downstream enrichment. Next, PCR amplification was performed to enrich the purified cDNA template. Finally, the cDNA libraries were sequenced using an Illumina HiSeq™ 2000. Three independent libraries as three biological replicates were sequenced for each sample. Transcript assembly and quantification of expression Considering the potential difference between diploid and octoploid Fragaria species, we first mapped the clean reads to the Fa genome (FAN_r1.1, http://strawberry-garden.kazusa.or.jp (August 8, 2018, date last accessed)); the unmapped reads were subsequently extracted and de novo assembled using the Trinity platform (Haas et al. 2013) with the parameters of ‘min_kmer_cov = 2, normalize_reads’. Considering the incompleteness of the Fa genome and the lack of annotation information, we chose Fv proteins (https://www.ncbi.nlm.nih.gov/genome (August 8, 2018, date last accessed)) for annotation. The pooled non-redundant transcripts were further annotated by alignment against protein databases including Swiss-Prot (http://www.uniprot.org (August 8, 2018, date last accessed)) and Uniref90 (https://www.uniprot.org/help/uniref (August 8, 2018, date last accessed)) using Diamond BLASTx (Buchfink et al. 2015). Transcript expression levels were quantified and normalized by TPM values using Salmon (Patro et al. 2017). Differential expression analysis was performed using the DESeq2 R package (Love et al. 2014). Transcripts with an absolute value of log2 FC ≥1 and adjusted P-value <0.05 were defined as significantly differentially expressed. qPCR validation The expression levels of selected transcripts involved in anthocyanin biosynthesis pathways were validated by qPCR using the same RNA samples as for sequencing. Specific primers were designed using beacon 7.0 software; cDNA was synthesized using a PrimeScript RT reagent Kit with gDNA Eraser (TAKARA). SYBR Green (TAKARA) was used for detection of the PCR products on a CFX96 Real-time reaction system (Bio-Rad). The FaActin gene (Accession: LC017712) was used as the internal control for normalization of gene expression. At least two well replicates were done for each sample, and three independent RNA samples of each plant sample were used as three biological replicates to ensure reproducibility and reliability. All primers used in this study are listed in Supplementary Table S11. Measurement of anthocyanins The anthocyanins were determined using HPLC, based on a previously described method (Donno et al. 2013). Briefly, 0.2 g frozen samples were finely ground and extracted in 2 ml of extraction solution (1% HCl in methanol) for 48 h at 4°C in darkness; extraction was repeated once. Samples were centrifuged at 13,000 r.p.m. for 15 min; the supernatants were combined and filtered using a 0.45 μm Nylon filter. A 10 μl aliquot of samples was subsequently injected into the HPLC system. Compound separations were achieved on a 250 mm×4.6 mm i.d., 5 μm reversed phase Silgreen ODS C18 column (Greenherbs Science and Technology), with 95% formic acid and methanol used as mobile phases. A linear gradient (95–0%) of formic acid in methanol was used for 20 min, followed by 100% methanol for 5 min. The column temperature was kept at 25°C; the flow rate was 1 ml min–1 and chromatograms were recorded at 510 nm. Anthocyanins were quantified by comparing them with external standards. Experiments were repeated three times with three independent samples referring to the three biological replicates. Transient overexpression of FaMYB10 The full-length cDNA sequence of FaMYB10 (Accession: EU155162) was amplified from strawberry ‘Xiaobai’ and cloned into the modified pCAMBIA1301 vector with the Cauliflower mosiac virus (CaMV) 35S promoter (Supplementary Fig. S4). Agrobacterium infiltration was performed based on the previously described method (Spolaore et al. 2001). Briefly, the Agrobacterium tumefaciens strain GV3101 containing the overexpression constructs was grown at 28°C in YEB medium containing 10 mM MES (pH 5.6) and appropriate antibiotics to reach a culture OD600 of approximately 0.8, the bacteria were then concentrated and re-suspended to a final OD600 of approximately 1.0. After incubation in MMA medium, the Agrobacterium suspension was injected into the whole fruits at the W stage, when they were still attached to the plants, by a sterile 1 ml syringe. The injected fruits were harvested 6–7 d (the day they turned fully red) after injection. As a control, fruits at the same stage were injected with bacteria containing an empty vector. At least three plants with at least three fruits on each were selected for infiltration; all the fruits at the W stage on each plant were infiltrated. Identification of putative lncRNAs and correlation analysis Identification of putative lncRNAs was performed according to a pipeline (Supplementary Fig. S5). Transcripts with class_code ‘u’ (unknown intergenic transcript), ‘o’ (generic exonic overlap with a reference transcript), ‘x’ (natural antisense transcript, NAT) and ‘i’ (intronic transcript) were extracted and subjected for size and open reading frame (ORF) selection. A perl script was used to extract transcripts longer than 200 nt and shorter than 100 amino acids. As we know, a real lncRNA does not have an ORF; therefore, the longest consecutive codon chain of the lncRNAs candidates was defined as the putative ORF. The coding potential capacities of the remaining transcripts were calculated by the coding potential calculator 2 (CPC2) program (Kang et al. 2017) and PLEK software (Li et al. 2014). Only transcripts with a coding potential score of less than –0.5 and which passed the coding potential calculation were retained. Subsequently, the retained transcripts were aligned to the Fv genome annotated protein sequences (https://www.ncbi.nlm.nih.gov (August 8, 2018, date last accessed)) using Diamond software (Buchfink et al. 2015). The transcripts with identity >90%, E-value <1.0E-10 and query coverage per HSP (qcovhsp) >80% were excluded. The remaining transcripts were then subjected to the Pfam database (https://pfam.xfam.org (August 8, 2018, date last accessed)), Uniref90 and nr database (http://www.ncbi.nlm.nih.gov (August 8, 2018, date last accessed)) to exclude the known protein domain-containing transcripts by the hmmscan tool (http://hmmer.org/ (August 8, 2018, date last accessed)) using default parameters (E-value <0.001). Moreover, to eliminate any known small RNAs, these remaining transcripts were blasted against the tRNA database (http://gtrnadb.ucsc.edu/ (August 8, 2018, date last accessed)), rRNA database (https://github.com/vaulot/pr2database (August 8, 2018, date last accessed)) and miRNA database (http://www.mirbase.org (August 8, 2018, date last accessed)) (E-value <0.001). The remaining transcripts were considered as putative lncRNAs. The co-expression of lncRNAs and mRNAs was calculated by R software using the Pearson correlation algorithm; the co-expressed pairs with correlation coefficient r > 0.8 and r < –0.8 at a P-value of <0.05 were recognized as significantly correlated pairs. Construction of lncRNA–miRNA–mRNA networks Known miRNAs in strawberry were collected from the literature (Ge et al. 2013, Li et al. 2013, Xu et al. 2013) and used as query for a homology search. The possibility of lncRNAs as the precursors of miRNAs was elucidated by SUmirFind and SUmirFold perl scripts (Lucas and Budak 2012). Potential miRNAs targets were predicted by the psRNAtarget online web server (http://plantgrn.noble.org/psRNATarget/ (August 8, 2018, date last accessed)) (Dai and Zhao 2011), with default parameters. The lncRNA–miRNA–mRNA interaction networks were visualized by Cytoscape 3.4.0 software (Shannon et al. 2003). Funding This work was supported by the Sichuan Doctoral Tutor Team Supported Project. Disclosures The authors have no conflicts of interest to declare. Acknowledgements We would like to thank the Institute of Pomology & Olericulture in Sichuan Agricultural University for providing the server to analyze the transcriptome data. References Afrin S. , Gasparrini M. , Forbes-Hernandez T.Y. , Reboredo-Rodriguez P. , Mezzetti B. , Varela-López A. , et al. ( 2016 ) Promising health benefits of the strawberry: a focus on clinical studies . J. Agric. Food Chem . 64 : 4435 – 4449 . Google Scholar Crossref Search ADS PubMed Aharoni A. , De Vos C. , Wein M. , Sun Z. , Greco R. , Kroon A. , et al. ( 2001 ) The strawberry FaMYB1 transcription factor suppresses anthocyanin and flavonol accumulation in transgenic tobacco . Plant J . 28 : 319 – 332 . Google Scholar Crossref Search ADS PubMed An J.P. , Qu F.J. , Yao J.F. , Wang X.N. , You C.X. , Wang X.F. , et al. ( 2017 ) The bZIP transcription factor MdHY5 regulates anthocyanin accumulation and nitrate assimilation in apple . Hortic. Res. 4 : 17023. Google Scholar Crossref Search ADS PubMed Andersen Ø.M. , Jordheim M. ( 2010 ) Chemistry of flavonoid-based colors in plants. In Comprehensive Natural Products II. Chemistry and Biology . Edited by Mander L. , Liu H.W. pp. 547 – 614 . Elsevier Press, Oxford . Baudry A. , Heim M.A. , Dubreucq B. , Caboche M. , Weisshaar B. , Lepiniec L. ( 2004 ) TT2, TT8, and TTG1 synergistically specify the expression of BANYULS and proanthocyanidin biosynthesis in Arabidopsis thaliana . Plant J . 39 : 366 – 380 . Google Scholar Crossref Search ADS PubMed Biswas S. , Hazra S. , Chattopadhyay S. ( 2016 ) Identification of conserved miRNAs and their putative target genes in Podophyllum hexandrum (Himalayan Mayapple) . Plant Gene 6 : 82 – 89 . Google Scholar Crossref Search ADS Buchfink B. , Xie C. , Huson D.H. ( 2015 ) Fast and sensitive protein alignment using DIAMOND . Nat. Methods 12 : 59. Google Scholar Crossref Search ADS PubMed Cagirici H.B. , Alptekin B. , Budak H. ( 2017 ) RNA sequencing and co-expressed long non-coding RNA in modern and wild wheats . Sci. Rep. 7 : 10670. Google Scholar Crossref Search ADS PubMed Carbone F. , Preuss A. , De Vos R.C. , D’Amico E. , Perrotta G. , Bovy A.G. , et al. ( 2009 ) Developmental, genetic and environmental factors affect the expression of flavonoid genes, enzymes and metabolites in strawberry fruits . Plant Cell Environ . 32 : 1117 – 1131 . Google Scholar Crossref Search ADS PubMed Celton J.M. , Gaillard S. , Bruneau M. , Pelletier S. , Aubourg S. , Martin-Magniette M.L ., et al. ( 2014 ) Widespread anti-sense transcription in apple is correlated with siRNA production and indicates a large potential for transcriptional and/or post-transcriptional control . New Phytol. 203 : 287 – 299 . Google Scholar Crossref Search ADS PubMed Chekanova J.A. ( 2015 ) Long non-coding RNAs and their functions in plants . Curr. Opin. Plant Biol. 27 : 207 – 216 . Google Scholar Crossref Search ADS PubMed Chen M. , Wang C. , Bao H. , Chen H. , Wang Y. ( 2016 ) Genome-wide identification and characterization of novel lncRNAs in Populus under nitrogen deficiency . Mol. Genet. Genomics 291 : 1663 – 1680 . Google Scholar Crossref Search ADS PubMed Chen Q. , Yu H.W. , Wang X.R. , Xie X.L. , Yue X.Y. , Tang H.R. ( 2012 ) An alternative cetyltrimethylammonium bromide-based protocol for RNA isolation from blackberry (Rubus L.) . Genet. Mol. Res. 11 : 1773 – 1782 . Google Scholar Crossref Search ADS PubMed Clark S.T. , Verwoerd W.S. ( 2011 ) A systems approach to identifying correlated gene targets for the loss of colour pigmentation in plants . BMC Bioinformatics 12 : 343. Google Scholar Crossref Search ADS PubMed Costantini L. , Malacarne G. , Lorenzi S. , Troggio M. , Mattivi F. , Moser C. , et al. ( 2015 ) New candidate genes for the fine regulation of the colour of grapes . J. Exp. Bot. 66 : 4427 – 4440 . Google Scholar Crossref Search ADS PubMed Dai X. , Zhao P.X. ( 2011 ) psRNATarget: a plant small RNA target analysis server . Nucleic Acids Res. 39 : W155 – W159 . Google Scholar Crossref Search ADS PubMed Dalmay T. ( 2013 ) Mechanism of miRNA-mediated repression of mRNA translation . Essays Biochem. 54 : 29. Google Scholar Crossref Search ADS PubMed Da Silva F.L. , Escribano-Bailón M.T. , Alonso J.J.P. , Rivas-Gonzalo J.C. , Santos-Buelga C. ( 2007 ) Anthocyanin pigments in strawberry . LWT-Food Sci. Technol . 40 : 374 – 382 . Google Scholar Crossref Search ADS De Lucia F. , Dean C. ( 2011 ) Long non-coding RNAs and chromatin regulation . Curr. Opin. Plant Biol. 14 : 168 – 173 . Google Scholar Crossref Search ADS PubMed De Vetten N. , Quattrocchio F. , Mol J. , Koes R. ( 1997 ) The an11 locus controlling flower pigmentation in petunia encodes a novel WD-repeat protein conserved in yeast, plants, and animals . Genes Dev . 11 : 1422 – 1434 . Google Scholar Crossref Search ADS PubMed Donno D. , Cavanna M. , Beccaro G. , Mellano M. , Torello-Marinoni D. , Cerutti A. , et al. ( 2013 ) Currants and strawberries as bioactive compound sources: determination of antioxidant profiles with HPLC-DAD/MS . J. Appl. Bot. Food Qual . 86 : 1 – 10 . Duan S. , Wang J. , Gao C. , Jin C. , Li D. , Peng D. , et al. ( 2018 ) Functional characterization of a heterologously expressed Brassica napus WRKY41-1 transcription factor in regulating anthocyanin biosynthesis in Arabidopsis thaliana . Plant Sci . 268 : 47 – 53 . Google Scholar Crossref Search ADS PubMed Franco-Zorrilla J.M. , Valli A. , Todesco M. , Mateos I. , Puga M.I. , Rubio-Somoza I. , et al. ( 2007 ) Target mimicry provides a new mechanism for regulation of microRNA activity . Nat. Genet. 39 : 1033. Google Scholar Crossref Search ADS PubMed Ge A. , Shangguan L. , Zhang X. , Dong Q. , Han J. , Liu H. , et al. ( 2013 ) Deep sequencing discovery of novel and conserved microRNAs in strawberry (Fragaria×ananassa) . Physiol. Plant. 148 : 387 – 396 . Google Scholar Crossref Search ADS PubMed Guan L. , Dai Z. , Wu B.-H. , Wu J. , Merlin I. , Hilbert G. , et al. ( 2016 ) Anthocyanin biosynthesis is differentially regulated by light in the skin and flesh of white-fleshed and teinturier grape berries . Planta 243 : 23 – 41 . Google Scholar Crossref Search ADS PubMed Gupta O.P. , Karkute S.G. , Banerjee S. , Meena N.L. , Dahuja A. ( 2017 ) Contemporary understanding of miRNA-based regulation of secondary metabolites biosynthesis in plants . Front. Plant Sci. 8 : 374. Google Scholar PubMed Haas B.J. , Papanicolaou A. , Yassour M. , Grabherr M. , Blood P.D. , Bowden J. , et al. ( 2013 ) De novo transcript sequence reconstruction from RNA-Seq: reference generation and analysis with Trinity . Nat. Protoc. 8 : 1494. Google Scholar Crossref Search ADS PubMed Han Y. , Vimolmangkang S. , Soria-Guerra R.E. , Rosales-Mendoza S. , Zheng D. , Lygin A.V. , et al. ( 2010 ) Ectopic expression of apple F3′H genes contributes to anthocyanin accumulation in the Arabidopsis tt7 mutant grown under nitrogen stress . Plant Physiol . 153 : 806 – 820 . Google Scholar Crossref Search ADS PubMed Härtl K. , Denton A. , Franz-Oberdorf K. , Hoffmann T. , Spornraft M. , Usadel B. , et al. ( 2017 ) Early metabolic and transcriptional variations in fruit of natural white-fruited Fragaria vesca genotypes . Sci. Rep. 7 : 45113. Google Scholar Crossref Search ADS PubMed Huang H. , Hu K. , Han K.T. , Xiang Q.Y. , Dai S.L. ( 2013 ) Flower colour modification of chrysanthemum by suppression of F3'H and overexpression of the exogenous Senecio cruentus F3'5'H gene . PLoS One 8 : e74395. Google Scholar Crossref Search ADS PubMed Jiang F. , Wang J.Y. , Jia H.F. , Jia W.S. , Wang H.Q. , Xiao M. ( 2013 ) RNAi-mediated silencing of the flavanone 3-hydroxylase gene and its effect on flavonoid biosynthesis in strawberry fruit . J. Plant Growth Regul. 32 : 182 – 190 . Google Scholar Crossref Search ADS Joshi R.K. , Megha S. , Basu U. , Rahman M.H. , Kav N.N.V. ( 2016 ) Genome wide identification and functional prediction of long non-coding RNAs responsive to Sclerotinia sclerotiorum infection in Brassica napus . PLoS One 11 : e0158784 . Google Scholar Crossref Search ADS PubMed Kang C. , Liu Z. ( 2015 ) Global identification and analysis of long non-coding RNAs in diploid strawberry Fragaria vesca during flower and fruit development . BMC Genomics 16 : 815. Google Scholar Crossref Search ADS PubMed Kang Y.J. , Yang D.C. , Kong L. , Hou M. , Meng Y.Q. , Wei L. , et al. ( 2017 ) CPC2: a fast and accurate coding potential calculator based on sequence intrinsic features . Nucleic Acids Res . 45 : W12 – W16 . Google Scholar Crossref Search ADS PubMed Kosar M. , Kafkas E. , Paydas S. , Baser K.H.C. ( 2004 ) Phenolic composition of strawberry genotypes at different maturation stages . J. Agric. Food Chem. 52 : 1586 – 1589 . Google Scholar Crossref Search ADS PubMed Li A. , Zhang J. , Zhou Z. ( 2014 ) PLEK: a tool for predicting long non-coding RNAs and messenger RNAs based on an improved k-mer scheme . BMC Bioinformatics 15 : 311. Google Scholar Crossref Search ADS PubMed Li H. , Mao W. , Liu W. , Dai H. , Liu Y. , Ma Y. , et al. ( 2013 ) Deep sequencing discovery of novel and conserved microRNAs in wild type and a white-flesh mutant strawberry . Planta 238 : 695 – 713 . Google Scholar Crossref Search ADS PubMed Li S. ( 2014 ) Transcriptional control of flavonoid biosynthesis: fine-tuning of the MYB–bHLH–WD40 (MBW) complex . Plant Signal. Behav. 9 : e27522. Google Scholar Crossref Search ADS PubMed Li Z. , Rana T.M. ( 2012 ) Molecular mechanisms of RNA-triggered gene silencing machineries . Acc. Chem. Res. 45 : 1122 – 1131 . Google Scholar Crossref Search ADS PubMed Liao Q. , Liu C. , Yuan X. , Kang S. , Miao R. , Xiao H. , et al. ( 2011 ) Large-scale prediction of long non-coding RNA functions in a coding–non-coding gene co-expression network . Nucleic Acids Res . 39 : 3864 – 3878 . Google Scholar Crossref Search ADS PubMed Lin X. , Xiao M. , Luo Y. , Wang J. , Wang H. ( 2013 ) The effect of RNAi-induced silencing of FaDFR on anthocyanin metabolism in strawberry (Fragaria×ananassa) fruit . Sci. Hortic . 160 : 123 – 128 . Google Scholar Crossref Search ADS Lin X.C. , Zhu Y. , Chen W.B. , Lin L.W. , Chen D.H. , Huang J.R. , et al. ( 2014 ) Integrated analysis of long non-coding RNAs and mRNA expression profiles reveals the potential role of lncRNAs in gastric cancer pathogenesis . Int. J. Oncol . 45 : 619 – 628 . Google Scholar Crossref Search ADS PubMed Lin-Wang K. , McGhie T.K. , Wang M. , Liu Y. , Warren B. , Storey R. , et al. ( 2014 ) Engineering the anthocyanin regulatory complex of strawberry (Fragaria vesca) . Front. Plant Sci. 5 : 651. Google Scholar Crossref Search ADS PubMed Liu J. , Jung C. , Xu J. , Wang H. , Deng S. , Bernad L. , et al. ( 2012 ) Genome-wide analysis uncovers regulation of long intergenic noncoding RNAs in Arabidopsis . Plant Cell 24 : 4333 – 4345 . Google Scholar Crossref Search ADS PubMed Liu J. , Wang H. , Chua N.H. ( 2015a ) Long noncoding RNA transcriptome of plants . Plant Biotechnol. J. 13 : 319 – 328 . Google Scholar Crossref Search ADS Liu X. , Hao L. , Li D. , Zhu L. , Hu S. ( 2015b ) Long non-coding RNAs and their biological roles in plants . Genomics Proteomics Bioinformatics 13 : 137 – 147 . Google Scholar Crossref Search ADS Liu Y. , Lin-Wang K. , Deng C. , Warran B. , Wang L. , Yu B. , et al. ( 2015c ) Comparative transcriptome analysis of white and purple potato to identify genes involved in anthocyanin biosynthesis . PLoS One 10 : e0129148. Google Scholar Crossref Search ADS Lloyd A. , Brockman A. , Aguirre L. , Campbell A. , Bean A. , Cantero A. , et al. ( 2017 ) Advances in the MYB–bHLH–WD repeat (MBW) pigment regulatory model: addition of a WRKY factor and co-option of an anthocyanin MYB for betalain regulation . Plant Cell Physiol . 58 : 1431 – 1441 . Google Scholar Crossref Search ADS PubMed Lopes-da-Silva F. , de Pascual-Teresa S. , Rivas-Gonzalo J. , Santos-Buelga C. ( 2002 ) Identification of anthocyanin pigments in strawberry (cv. Camarosa) by LC using DAD and ESI-MS detection . Eur. Food Res. Technol. 214 : 248 – 253 . Google Scholar Crossref Search ADS Love M.I. , Huber W. , Anders S. ( 2014 ) Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2 . Genome Biol. 15 : 550. Google Scholar Crossref Search ADS PubMed Lucas S.J. , Budak H. ( 2012 ) Sorting the wheat from the chaff: identifying miRNAs in genomic survey sequences of Triticum aestivum chromosome 1AL . PLoS One 7 : e40859 . Google Scholar Crossref Search ADS PubMed Ma H. , Hao Y. , Dong X. , Gong Q. , Chen J. , Zhang J. , et al. ( 2012 ) Molecular mechanisms and function prediction of long noncoding RNA . ScientificWorldJournal 2012 : 541786 Google Scholar PubMed Mazza G. , Miniati E. ( 1993 ) Anthocyanins in Fruits, Vegetables, and Grains . CRC Press , Boca Raton, FL . Medina-Puche L. , Cumplido-Laso G. , Amil-Ruiz F. , Hoffmann T. , Ring L. , Rodríguez-Franco A. , et al. ( 2014 ) MYB10 plays a major role in the regulation of flavonoid/phenylpropanoid metabolism during ripening of Fragaria×ananassa fruits . J. Exp. Bot. 65 : 401 – 417 . Google Scholar Crossref Search ADS PubMed Mochizuki Y. , Iwasaki Y. , Fuke M. , Ogiwara I. ( 2014 ) Analysis of a high-yielding strawberry (Fragaria×ananassa Duch.) cultivar ‘Benihoppe’ with focus on root dry matter and activity . J. Jpn. Soc. Hort. Sci. 83 : 142 – 148 . Google Scholar Crossref Search ADS Nakatsuka T. , Yamada E. , Saito M. , Fujita K. , Nishihara M. ( 2013 ) Heterologous expression of gentian MYB1R transcription factors suppresses anthocyanin pigmentation in tobacco flowers . Plant Cell Rep. 32 : 1925 – 1937 . Google Scholar Crossref Search ADS PubMed Nie M. , Deng Z.-L. , Liu J. , Wang D.Z. ( 2015 ) Noncoding RNAs, emerging regulators of skeletal muscle development and diseases . BioMed Res. Int . 2015 : 676575. Google Scholar PubMed Nithin C. , Thomas A. , Basak J. , Bahadur R.P. ( 2017 ) Genome-wide identification of miRNAs and lncRNAs in Cajanus cajan . BMC Genomics 18 : 878 . Google Scholar Crossref Search ADS PubMed Patro R. , Duggal G. , Love M.I. , Irizarry R.A. , Kingsford C. ( 2017 ) Salmon provides fast and bias-aware quantification of transcript expression . Nat. Methods 14 : 417 – 419 . Google Scholar Crossref Search ADS PubMed Pillet J. , Folta K.M. ( 2015 ) Pigments in strawberry. In Pigments in Fruits and Vegetables . Edited by Chen C. pp. 205 – 216 . Springer Press , New York . Reddy A.M. , Reddy V.S. , Scheffler B.E. , Wienand U. , Reddy A.R. ( 2007 ) Novel transgenic rice overexpressing anthocyanidin synthase accumulates a mixture of flavonoids leading to an increased antioxidant potential . Metab. Eng . 9 : 95 – 111 . Google Scholar Crossref Search ADS PubMed Salvatierra A. , Pimentel P. , Moya-Leon M.A. , Caligari P.D. , Herrera R. ( 2010 ) Comparison of transcriptional profiles of flavonoid genes and anthocyanin contents during fruit development of two botanical forms of Fragaria chiloensis ssp. chiloensis . Phytochemistry 71 : 1839 – 1847 . Google Scholar Crossref Search ADS PubMed Schaart J.G. , Dubos C. , Romero De La Fuente I. , Houwelingen A.M. , Vos R.C. , Jonker H.H. , et al. ( 2013 ) Identification and characterization of MYB–bHLH–WD40 regulatory complexes controlling proanthocyanidin biosynthesis in strawberry (Fragaria×ananassa) fruits . New Phytol. 197 : 454 – 467 . Google Scholar Crossref Search ADS PubMed Shafiq S. , Li J. , Sun Q. ( 2016 ) Functions of plants long non-coding RNAs . Biochim. Biophys. Acta. 1859 : 155. Google Scholar Crossref Search ADS PubMed Shannon P. , Markiel A. , Ozier O. , Baliga N.S. , Wang J.T. , Ramage D. , et al. ( 2003 ) Cytoscape: a software environment for integrated models of biomolecular interaction networks . Genome Res . 13 : 2498 – 2504 . Google Scholar Crossref Search ADS PubMed Sharma S. , Taneja M. , Tyagi S. , Singh K. , Upadhyay S.K. ( 2017 ) Survey of high throughput RNA-seq data reveals potential roles for lncRNAs during development and stress response in bread wheat . Front. Plant Sci . 8 : 1019 . Google Scholar Crossref Search ADS PubMed Shin D.H. , Choi M.-G. , Kang C.S. , Park C.-S. , Choi S.B. , Park Y.I. ( 2016 ) Overexpressing the wheat dihydroflavonol 4-reductase gene TaDFR increases anthocyanin accumulation in an Arabidopsis dfr mutant . Genes Genom. 38 : 333 – 340 . Google Scholar Crossref Search ADS Sompornpailin K. , Makita Y. , Yamazaki M. , Saito K. ( 2002 ) A WD-repeat-containing putative regulatory protein in anthocyanin biosynthesis in Perilla frutescens . Plant Mol. Biol . 50 : 485 – 495 . Google Scholar Crossref Search ADS PubMed Sondheimer E. , Karash C.B. ( 1956 ) The major anthocyanin pigments of the wild strawberry (Fragaria vesca) . Nature 178 : 648 – 649 . Google Scholar Crossref Search ADS Song D. , Yang Y. , Yu B. , Zheng B. , Deng Z. , Lu B.L. , et al. ( 2009 ) Computational prediction of novel non-coding RNAs in Arabidopsis thaliana. BMC Bioinformatics 10 : S36 . Google Scholar Crossref Search ADS PubMed Spolaore S. , Trainotti L. , Casadoro G. ( 2001 ) A simple protocol for transient gene expression in ripe fleshy fruit mediated by agrobacterium . J. Exp. Bot. 52 : 845 – 850 . Google Scholar Crossref Search ADS PubMed Verdier J. , Zhao J. , Torres-Jerez I. , Ge S. , Liu C. , He X. , et al. ( 2012 ) MtPAR MYB transcription factor acts as an on switch for proanthocyanidin biosynthesis in Medicago truncatula . Proc. Natl. Acad. Sci. USA 109 : 1766 – 1771 . Google Scholar Crossref Search ADS Wang J. , Meng X. , Dobrovolskaya O.B. , Orlov Y.L. , Chen M. ( 2017 ) Non-coding RNAs and their roles in stress response in plants . Genomics Proteomics Bioinformatics 15 : 301 – 312 . Google Scholar Crossref Search ADS PubMed Wang K.C. , Chang H.Y. ( 2011 ) Molecular mechanisms of long noncoding RNAs . Mol. Cell. 43 : 904 – 914 . Google Scholar Crossref Search ADS PubMed Wang S.Y. , Zheng W. , Galletta G.J. ( 2002 ) Cultural system affects fruit quality and antioxidant capacity in strawberries . J. Agric. Food Chem. 50 : 6534 – 6542 . Google Scholar Crossref Search ADS PubMed Wang Y. , Wang X. , Deng W. , Fan X. , Liu T.-T. , He G. , et al. ( 2014 ) Genomic features and regulatory roles of intermediate-sized non-coding RNAs in Arabidopsis . Mol. Plant. 7 : 514 – 527 . Google Scholar Crossref Search ADS PubMed Wu X. , Gong Q. , Ni X. , Zhou Y. , Gao Z. ( 2017 ) UFGT: the key enzyme associated with the petals variegation in Japanese apricot . Front. Plant Sci. 8 : 108. Google Scholar PubMed Wu X. , Prior R.L. ( 2005 ) Systematic identification and characterization of anthocyanins by HPLC-ESI-MS/MS in common foods in the United States: fruits and berries . J. Agric. Food Chem. 53 : 2589 – 2599 . Google Scholar Crossref Search ADS PubMed Xie S. , Song C. , Wang X. , Liu M. , Zhang Z. , Xi Z. ( 2015 ) Tissue-specific expression analysis of anthocyanin biosynthetic genes in white- and red-fleshed grape cultivars . Molecules 20 : 22767 – 22780 . Google Scholar Crossref Search ADS PubMed Xin M. , Wang Y. , Yao Y. , Song N. , Hu Z. , Qin D. , et al. ( 2011 ) Identification and characterization of wheat long non-protein coding RNAs responsive to powdery mildew infection and heat stress by using microarray analysis and SBS sequencing . BMC Plant Biol. 11 : 61. Google Scholar Crossref Search ADS PubMed Xu W. , Dubos C. , Lepiniec L. ( 2015 ) Transcriptional control of flavonoid biosynthesis by MYB–bHLH–WDR complexes . Trends Plant Sci. 20 : 176 – 185 . Google Scholar Crossref Search ADS PubMed Xu X. , Yin L. , Ying Q. , Song H. , Xue D. , Lai T. , et al. ( 2013 ) High-throughput sequencing and degradome analysis identify miRNAs and their targets involved in fruit senescence of Fragaria ananassa . PLoS One 8 : e70959. Google Scholar Crossref Search ADS PubMed Yoon J.H. , Abdelmohsen K. , Gorospe M. ( 2014 ) Functional interactions among microRNAs and long noncoding RNAs . Semin. Cell Dev. Biol. 34 : 9 – 14 . Google Scholar Crossref Search ADS PubMed Zhan S. , Dong Y. , Zhao W. , Guo J. , Zhong T. , Wang L. , et al. ( 2016 ) Genome-wide identification and characterization of long non-coding RNAs in developmental skeletal muscle of fetal goat . BMC Genomics 17 : 666. Google Scholar Crossref Search ADS PubMed Zhang Y. , Li W. , Dou Y. , Zhang J. , Jiang G. , Miao L. , et al. ( 2015 ) Transcript quantification by RNA-Seq reveals differentially expressed genes in the red and yellow fruits of Fragaria vesca . PLoS One 10 : e0144356. Google Scholar Crossref Search ADS PubMed Zhu B. , Yang Y. , Li R. , Fu D. , Wen L. , Luo Y. , et al. ( 2015 ) RNA sequencing and functional analysis implicate the regulatory role of long non-coding RNAs in tomato fruit ripening . J. Exp. Bot. 66 : 4483 – 4495 . Google Scholar Crossref Search ADS PubMed Zufall R.A. , Rausher M.D. ( 2004 ) Genetic changes associated with floral adaptation restrict future evolutionary potential . Nature 428 : 847 – 850 . Google Scholar Crossref Search ADS PubMed Abbreviations Abbreviations ANR anthocyanidin reductase ANS anthocyanidin synthase bHLH basic helix–loop–helix bZIP basic leucine zipper CHI chalcone isomerase CHS chalcone synthase CPC2 coding potential calculator 2 DFR dihydroflavonol 4-reductase DPA days post-anthesis ERF ethylene-responsive factor Fa Fragaria×ananassa FC fold change F3H flavanone 3-hydroxylase F3'H flavonoid 3'-hydroxylase FLS flavonol synthase Fv Fragaria vesca G stage green stage LAR leucoanthocyanidin reductase lncRNA long non-coding RNA MBW MYB–bHLH–WD40 miRNA microRNA ncRNA non-coding RNA ORF open reading frame qPCR real-time quantitative PCR R stage full red stage RNAi RNA interference RNAseq RNA sequencing T stage turning stage TF transcription factor TPM transcripts per kilobase million UFGT anthocyanidin 3-O-glucosylthransferase W stage white stage © The Author(s) 2018. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Plant and Cell PhysiologyOxford University Press

Published: Sep 1, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off