The light-induced transcription factor FtMYB116 promotes accumulation of rutin in Fagopyrum tataricum

2.1 Plant material and growth conditions

Tartary buckwheat seeds were soaked in water for 20 min, after which the seed coats were peeled off. Peeled seeds were initially washed with 75% ethanol for 1 min and then soaked in 0.5% mercury for 8 min, followed by six 1-min washes with sterile water. Thereafter, the seeds were blotted on filter paper to remove excess moisture and then sown onto Murashige and Skoog (MS) medium containing 1% sucrose and 0.8% agar and were incubated in darkness for 4 days at 25°C, followed by various continuous light treatments for 2 days. Plants were illuminated with light-emitting diodes (LEDs) emitting the following wavelengths: red light (670 nm), blue light (470 nm), and far-red light (735 nm). For all treatments, the light intensity was 50 ± 5 μmol m⁻² s⁻¹, and each treatment comprised three biological replicates. Following these light treatments, plant samples were ground into powder and stored in liquid nitrogen. The flavonoid content in these preparations was determined using a 0.1-g amount of the powder, whereas a further 0.1-g sample was used to extract RNA for sequencing.

2.2 RNA sequencing

Plant total RNA was isolated using an RNA Extraction Kit (Tiangen, Beijing, China) according to the manufacturer's instructions. RNA quality was assessed using a Bioanalyzer 2100 instrument (Agilent, USA), and the RNA integrity number of all RNA samples was found to be >7.0. Sequencing libraries were prepared according to the manufacturer's instructions (Illumina, USA), and RNA sequencing was performed using an Illumina X Ten sequencing system (Illumina Inc., San Diego, CA, USA).

2.3 RNA-seq data analysis

After removing Illumina adapter sequences and trimming low-quality bases, approximately 46.9 G bases of clean data were used in our analysis. These reads were subsequently mapped to the tartary buckwheat reference genome (L. Zhang, Li, et al., 2017) using STAR v2.5 software. Differentially expressed genes (DEGs) were screened based on their FPKM (reads per kilobase of exon per million mapped reads) values (Mortazavi, Williams, Mccue, Schaeffer, & Wold, 2008). FPKM values were calculated using Htseq software. For DEG analysis, we used DESeq2 software (log2FoldChange ≥ 1).

2.4 Transcription factor identification and coexpression analysis

Transcription factor families were identified using hmmscan software. To select the possible transcription factors regulating genes involved in flavonoid biosynthesis, four highly expressed genes, F3'H (FtPinG0002353900), flavone-3-hydroxylase (F3H; FtPinG0006662600), 4-coumarate:CoA ligase (4CL; FtPinG0005072700), and cinnamate-4-hydroxylase (C4H; FtPinG0001575100), were used in coexpression analysis with a default value 0.05 (Ariani & Gepts, 2015). Paired genes showing a Pearson correlation coefficient (r) greater than 0.95 were considered as significantly coexpressed and selected to build a coexpression network using the Perl script (T. Zhang, Song, et al., 2017).

2.5 Quantitative RT-PCR

Plant total RNA was isolated using an RNA Extraction Kit (Tiangen), and first-strand cDNA was synthesized using oligo d(T)18 by reverse transcriptase (Invitrogen, Carlsbad, CA, USA) and then diluted (1:100) for use in quantitative reverse transcription polymerase chain reaction (qRT-PCR) with SYBR Premix Ex Taq Mix (Takara, Dalian, Liaoning, China) in a total volume of 20 μl. Real-time PCR was performed using a LightCycler 480 System (Roche, Basel, Switzerland), following the manufacturer's instructions. Three biological replicates were used for each sample, and the expression level was normalized to that of actin, which was used as the control gene.

2.6 Yeast hybrid assays

The AD-FtMYB116 effectors were cotransformed with the F3′Hp:LacZ reporters into yeast strain EGY48, and transformants were selected and grown on SD/-Trp-Ura dropout media. The transformants were further grown on SD/-Trp-Ura dropout media containing 20-mg ml⁻¹ X-gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside) for colour development.

2.7 Electrophoretic mobility shift assay

The MBP-FtMYB116 recombinant proteins were expressed and purified from Escherichia coli. Electrophoretic mobility shift assay experiments were performed as previously described (Z. Jiang, Xu, Jing, Tang, & Lin, 2016).

2.8 Transient luciferase expression assays

The F3′Hp:LUC reporter plasmid and FtMYB116 effector construct were transformed into Agrobacterium. The Agrobacterium suspension was slowly injected into the whole leaves of N. tabacum. After injection, the recipient N. tabacum plants were initially left in the dark overnight and thereafter grown under light/dark cycle of 16/8 hr. At 48 hr after injection, LUC fluorescence in leaf blades was observed using NightSHADE LB 985 (Berthold, Germany).

2.9 Transformation of the hairy roots of tartary buckwheat

The hypocotyls and cotyledons of 7- to 10-day tartary buckwheat seedlings were selected as explants. Cotyledons were cut into small squares, and hypocotyls were cut into approximately 0.5-cm segments, and these samples were precultured on MS solid medium for 1 day. A suspension of bacteria harbouring exogenous gene plasmids was poured into the triangular bottles containing the leaf and hypocotyl explants for 10 min and then poured out. Thereafter, the explants were transferred to sterile triangular bottles and cocultured with a bacterial solution for 3 days in the dark. The explants were rinsed with sterile water 4–5 times until the water was clear and were then placed on MS medium. The appearance of hairy roots was observed 2 weeks later. Those hairy roots showing rapid growth were selected, cut into 2- to 3-cm pieces, and propagated in liquid MS medium (containing 100-mg L⁻¹ carbenicillin) at 25°C on a shaking table rotating at 80 rpm.

2.10 Sample treatment for secondary metabolite analysis

Fresh seedlings were ground into powder, 0.1 g of which was extracted overnight at 4°C with 1.0-ml pure methanol (or 70% aqueous methanol) containing 0.1-mg L⁻¹ lidocaine. Lipid-soluble extracts were absorbed, and 0.4 ml of each extract was mixed and filtrated (SCAA-104; 0.22-μm pore size) before liquid chromatography–mass spectrometry (LC-MS) analysis.

2.11 LC-MS analysis

The analytical conditions used for LC-MS were as follows: column, Agilent Eclipse Plus C18 column (250 mm × 4.6 mm, 5 μm; Santa Clara, CA, USA); solvent system, water (0.1% formic acid) and acetonitrile (0.1% formic acid); gradient programme, 5:95 v/v, from 0 to 12 min and 95:5 v/v from 12 to 15 min; flow rate, 0.2 ml min⁻¹; temperature, 45°C; and injection volume, 1 μl.

3.1 Morphology and flavonoid content of tartary buckwheat during sprout development under far-red, red, and blue light

Tartary buckwheat seedlings were initially grown in the dark for 4 days and then subjected to various light treatments for a further 2 days. The hypocotyls of seedlings in dark and under LED red light were slightly longer than those under LED far-red and blue light. Further, in the dark, the cotyledons of seedlings were closed and yellow in colour, whereas the hypocotyl was pale white. Under LED red light, the cotyledons of seedlings were slightly open and deep green in colour and the hypocotyl was light red. Under LED far-red light, the cotyledons of seedlings were wide open and light green in colour and the hypocotyl was light red. Under LED blue light, the cotyledons of seedlings were closed and deep green in colour and the hypocotyl was deep red (Figure 1a). The different morphology of seedlings indicated that the various wavelengths of light have different effects on tartary buckwheat seedlings.

Using LC-MS analysis, we analysed the content of 28 compounds (Table S1). Compared with the dark control, the provision of far-red, red, and blue light all promoted increases in rutin, astragalin, and quercitrin content, with blue light having the most pronounced effect, followed by red and then far-red light (Figure 1b). Although we found that red and blue light promoted increases in the content of quercetin, isorhamnetin-3-O-glucoside, isorhamnetin, and trifolin, far-red light had no effect. Again, the effect of blue light was observed to be stronger than that of red light (Figure 1b). We also identified the presence of orientin, vitexin, isovitexin, and catechin-O-β-d-glucoside under various wavelengths of light, but no obvious changes were observed (Figure S1). These results showed that light can promote increases in the content of flavonoids in buckwheat seedlings and that blue light is the most effective in this regard, followed by red light and then far-red light.

3.2 Comparative analysis of transcriptional profiles

Samples of plants illuminated under different light treatments were subjected to Illumina X Ten paired-end sequencing. The sequencing data have been deposited in the National Center for Biotechnology Information database (accession number: SRP157461). To identify DEGs in the different treatment groups, fragments per kilobase of exon per million mapped reads (FPKM) values were calculated; complete data can be found in Table S2. The DEGs are listed in Table S3. We found that under red, blue, and far-red light treatments, 2,580, 4,113, and 3,775 genes, accounting for 7.7%, 12.3%, and 11.3% of the total genes (33,366), respectively, were significantly up-regulated by at least two-fold (Figure 2a). Conversely, under red, blue, and far-red light, 2,197, 3,411, and 4,791 genes, accounting for 6.6%, 10.2%, and 14.4% of the total genes, respectively, were significantly down-regulated by at least two-fold (Figure 2b). Venn diagrams constructed for the number of up-regulated and down-regulated genes among different light treatments display overlapping and uniquely expressed genes under each condition. Far-red light had the most prominent influence on gene expression, resulting in the up-regulation and down-regulation of 11.3% and 14.4% of DEGs, respectively, whereas red light had the least effect, leading to up-regulation and down-regulation of only 7.7% and 6.6% of the total DEGs, respectively. Overall, among the three treatment groups, 1,004 and 902 genes were commonly up-regulated or down-regulated between at least two treatments (Figure 2).

3.3 Genes related to flavonoid biosynthesis in tartary buckwheat

We determined the expression of genes involved in flavonoid synthesis pathways according to our annotation of the tartary buckwheat genome (Figure S2). Through the phenylpropanoid pathway, flavonoids are synthesized from several individual branches and can be classified into different groups of derivatives (Figure 3). Phenylalanine is converted to 4-coumaroyl-CoA via the catalytic action of PAL, C4H, and 4CL (Dixon & Lamb, 1979; Russell, 1971). Subsequently, chalcone synthase (CHS) catalyses the conversion of 4-coumaroyl-CoA to naringenin chalcone (Koes, Spelt, van den Elzen, & Mol, 1989). The CHS gene was found to be induced most strongly by red light. Subsequently, CHI catalyses the conversion of naringenin chalcone to naringenin (Mona & Christopher, 1987), which is then converted to flavones, such as orientin and vitexin, through CGT catalysis (Nagatomo et al., 2014). Naringenin can also be converted to dihydrokaempferol and dihydroquercetin via F3H and F3′H, respectively (Cheng, Han, Wu, & Lou, 2014; Hagmann, Heller, & Grisebach, 1983). At this point, the pathway divides into two branches. In one branch, dihydroquercetin is converted to proanthocyanidin through the catalysis of dihydroflavonol reductase (DFR), whereas in the other branch, flavonol synthase (FLS) converts dihydroquercetin to quercetin, which subsequently is converted to rutin by F3G6″RhaT (Holton, Brugliera, & Tanaka, 1993; Koja et al., 2018). The expression of CHI, F3H, and FLS was induced by red and blue light, whereas F3′H was specifically induced by blue light. In contrast, the expression of DFR was inhibited by light.

3.4 Validation of differential expression via real-time qRT-PCR

The expression levels of differentially regulated genes were validated using qRT-PCR. For the selected genes, the qRT-PCR expression profile showed positive correlations with the transcriptome data (Figure 4) and the primers used in the qRT-PCR were listed in Table S4. PAL and CHS were strongly induced, whereas C4H and DFR were inhibited by far-red and blue light, and red light had weak effects on all these genes. Furthermore, CHI, F3H, F3′H, and FLS were strongly induced by red and blue light, with the effect of blue light being more obvious than that of red light. None of these genes were induced by far-red light.

3.5 Identification of coexpressed transcription factors that directly regulate the synthesis of flavonoids

To identify the specific transcription factors that may play roles in regulating flavonoid biosynthesis gene expression, we performed coexpression analysis and results are presented in Table S5. Through this analysis, we identified 10, 33, 29, and 21 transcription factors that were coexpressed with F3′H, F3H, 4CL, and 4CH, respectively (Figure 5 and Table S6). Among these transcription factors, we found that one MYB transcription factor (FtPinG0007711300.01), which we named FtMYB116, was coexpressed with all four flavonoid biosynthesis genes (Figure 5). We accordingly focused subsequent efforts on FtMYB116 as the transcription factor most likely to directly regulate the synthesis of flavonoids.

3.6 The expression pattern of FtMYB116

To further characterize FtMYB116, we determined the expression pattern of this transcription factor in different tissues and under different light conditions. The results showed that, compared with darkness, red and blue light significantly promoted the expression of FtMYB116, whereas far-red light was virtually ineffective in this regard. We observed that the expression of FtMYB116 was highest in leaves, followed by seeds, flowers, and stems, with the lowest expression being detected in the roots (Figure 6).

3.7 FtMYB116 directly activates the expression of F3′H

To study the direct effect of FtMYB116 on rutin synthesis, we performed yeast one-hybrid assays. Fragments (2,000 bp) of the promoters of CHS, CHI, F3H, F3′H, and FLS were cloned into pLacZ-2μ reporter vectors, and FtMYB116 was cloned into pB42AD (GAL4 activation domain [AD] fused). As shown in Figure 7b, FtMYB116 bound to the promoter of F3′H, but not to the promoters of the other four examined genes (data not shown). Subsequent bioinformatic analysis of the promoter of F3'H indicated that there are three potential motifs (MBS, MBSI, and MRE) to which FtMYB116 could bind (Figure S3). Because MRE is light induced and Hartmann has previously reported that MYB transcription factors can bind to the MRE (TACCTACC) motif (Hartmann, Sagasser, Mehrtens, Stracke, & Weisshaar, 2005), we conducted a further detailed analysis of MRE in yeast. As shown in Figure 7a,b, when TACCTACC was mutated to GCAAGCAA, FTMYB116 could not bind to the mutated F3′H promoter in yeast.

To verify the direct interaction between FtMYB116 and the F3′H promoter in vitro, we expressed MBP-FtMYB116 (fused with the maltose-binding protein) recombinant protein and performed an electrophoresis mobility shift assay. As shown in Figure 7c, the MBP-FtMYB116 fusion protein promotes a shift in the band of the biotin-labelled DNA fragments containing MRE motifs. The intensity of this shifted band was decreased with the addition of excess levels of unlabelled wild-type fragments.

A transient luciferase activity assay was performed in N. tabacum to further examine the regulation of F3′H by MYB116. As shown in Figure 7d, no luciferase fluorescence was observed in two leaves of N. tabacum injected with water (control). When leaves were injected with Agrobacterium transformed only with the F3′H promoter, there was background fluorescence. However, when leaves were injected with Agrobacterium cotransformed with the F3′H promoter and FtMYB116, there was a significant increase in the fluorescence value of leaves, indicating that FtMYB116 can activate F3′H expression.

3.8 Overexpression of FtMYB116 promotes rutin biosynthesis in tartary buckwheat hairy roots

Transcriptome data revealed that FtMYB116 induces the expression of CHI, FLS, and F3H, and thus, FtMYB116 may promote the accumulation of flavonoids. To verify whether FtMYB116 affects the synthesis of flavonoids, we generated transgenic hairy roots of tartary buckwheat overexpressing FtMYB116. Empty vector was also introduced into hairy roots as a control. Three independent transgenic hairy root lines expressing FtMYB116 were used for further functional analysis, in which we investigated the expression levels of FtMYB116 (Figure S4). The expression of FtMYB116 in these lines was significantly higher than that in the control, indicating that FtMYB116 was expressed in hairy roots. Subsequently, we determined the content of flavonoids in the three transgenic lines. As shown in Figure 8, the levels of rutin and quercetin were significantly higher in these lines than in control plants containing an empty vector, indicating that FtMYB116 promotes flavonoid biosynthesis. Subsequently, qRT-PCR was performed to investigate the expression patterns of five rutin biosynthesis genes (PAL, F3H, FLS, CHS, and F3′H) in tartary buckwheat hairy root lines. The transcripts of all the genes were significantly up-regulated in the FtMYB116-overexpressing lines (Figure 8c).