Genome-Wide Identification of Direct Targets of the TTG1–bHLH–MYB Complex in Regulating Trichome Formation and Flavonoid Accumulation in Arabidopsis Thaliana

Extensive studies have shown that the MBW complex consisting of three kinds of regulatory proteins, MYB and basic helix–loop–helix (bHLH) transcription factors and a WD40 repeat protein, TRANSPARENT TESTA GLABRA1 (TTG1), acts in concert to promote trichome formation and flavonoid accumulation in Arabidopsis thaliana. TTG1 functions as an essential activator in these two biological processes. However, direct downstream targets of the TTG1-dependent MBW complex have not yet been obtained in the two biological processes at the genome-wide level in A. thaliana. In the present study, we found, through RNA sequencing and quantitative real-time PCR analysis, that a great number of regulatory and structural genes involved in both trichome formation and flavonoid accumulation are significantly downregulated in the young shoots and expanding true leaves of ttg1-13 plants. Post-translational activation of a TTG1-glucocorticoid receptor fusion protein and chromatin immunoprecipitation assays demonstrated that these downregulated genes are directly or indirectly targeted by the TTG1-dependent MBW complex in vivo during trichome formation and flavonoid accumulation. These findings further extend our understanding of the role of TTG1-dependent MBW complex in the regulation of trichome formation and flavonoid accumulation in A. thaliana.


Introduction
TRANSPARENT TESTA GLABRA1 (TTG1) encodes a WD40 repeat transcription factor that plays pleiotropic roles in the regulation of seed development and postembryonic processes in Arabidopsis thaliana [1][2][3][4][5][6][7][8]. During seed development, it not only promotes the biosynthesis of proanthocyanidins (PAs) [1][2][3]8,9] but also accelerates the production of mucilage and columella [1,8,10,11]. By contrast, it negatively regulates the accumulation of seed storage reserves, including fatty acids and proteins during seed maturation [7,8]. In the postembryonic process, TTG1 functions in root development including root length and hairs [12,13] and the response to abiotic stresses [13,14]. Further, TTG1 also acts as a key transcriptional activator in trichome formation [1,3,12,15] and flavonoid deposition [1][2][3]9,16], The loss-of-function mutant of ttg1-13 (CS67772) as a mixed Columbia/Ler background was obtained by fast neutron mutagenesis from the Arabidopsis Biological Resource Center (ABRC). This mutant was backcrossed twice to the wild type Col-0 to purify its background and eliminate other possible mutations [14], and then used in this study. There are no trichome formation and flavonoid deposition in ttg1-13 young shoots and true leaves ( Figure S1, [14]), which is consistent with the previous evidence that TTG1 positively promotes trichome formation and flavonoid accumulation in A. thaliana [1][2][3]9,12,15,16]. Trichomes are commonly present in the surface of stems and leaves, and anthocyanins, the most conspicuous class of flavonoids, are frequently produced in young shoots and expanding leaves [47]. Further, TTG1 is highly expressed in young shoots and expanding true leaves [25,46]. Therefore, to identify downstream targeted genes of the TTG1-dependent MBW complex involved in trichome formation and flavonoid accumulation at the genome-wide level, the tissues GSTF12 (AT5G17220) −3.73 Involved in transport and accumulation of both anthocyanins and proanthocyanidins [81,82]

Validation of Downstream Targets Related to Trichome Formation and Flavonoid Accumulation in ttg1-13 Young Shoots and Expanding True Leaves
To further confirm the expression of these downregulated genes related to trichome formation and flavonoid accumulation, quantitative real-time PCR (qRT-PCR) was then used to analyze their dynamic expression levels in the ttg1-13 young shoots and expanding true leaves at 18, 20, and 22 DAG. The results showed that the expression levels of all these genes involved in trichome formation and flavonoid accumulation are significantly reduced at 20 DAG, which is highly consistent with the RNA-seq results (Figures 1 and 2; Tables 1 and 2, Tables S1 and S2). For the trichome-inducing genes, the expression levels of 9 regulatory genes, SQUAMOSA PROMOTER BINDING PROTEIN-LIKE 8 (SPL8), ENHANCER OF TRYAND CPC 1 (ETC1), BRANCHLESS TRICHOMES (BLT), HOMEODOMAIN GLABROUS11 (HDG11), CPC, MYB106, MYB5, SIAMESE (SIM), and MYB23, and 1 structural gene, SMALLER WITH VARIABLE BRANCHES (SVB), were always significantly lower in the ttg1-13 young shoots and expanding true leaves relative to the wild type plants at 18 and 22 DAG (Figure 1; Table 1 and Table S1). On the other hand, the expression levels of 11 genes related to flavonoid biosynthesis were also greatly downregulated in the ttg1-13 young shoots and expanding true leaves at 18 and 22 DAG, which includes 1 regulatory gene, ANTHOCYANINLESS2 (ANL2), and 10 structural genes in the flavonoid biosynthetic pathway, PHENYLALANINE AMMONIA-LYASE4 (PAL4),  Table 2 and Table S1). It is notable that 4 regulatory genes involved in both trichome formation and flavonoid accumulation, namely MYC2, GL2, TTG2, and TT8, were dramatically downregulated in the ttg1-13 young shoots and expanding true leaves at 18 and 22 DAG (Figures 1 and 2; Tables 1 and 2, and Table S1). Previous studies have demonstrated that TTG2, TT8, DFR, F3'H, and UGT79B1 are targets of TTG1 [25,44,69], which is a proof of our successful gene expression analysis in this study. Taken together, these results further indicated that the TTG1-dependent MBW complex promotes trichome formation and flavonoid accumulation by activating a series of regulatory and structural genes involved in trichome development and flavonoid biosynthesis, respectively, in A. thaliana young shoots and expanding true leaves.

Identification of Direct Downstream Targets Contributing to Trichome Formation and Flavonoid Accumulation Regulated by the TTG1-Dependent MBW Complex in Young Shoots and Expanding True Leaves
To investigate how the TTG1-dependent MBW complex controls the mRNA expression of downstream targeted genes, we created a steroid-inducible version of TTG1 in the background of ttg1-13, in which the TTG1 gene was fused to the rat glucocorticoid receptor (GR) and driven by the 35S promoter. We isolated a ttg1-13 35S:TTG1-GR transgenic line, which fully rescued the phenotypes of ttg1-13 leaves without trichomes and flavonoids after DEX treatment every other day after germination ( Figure S2), whereas the mock-treated ttg1-13 35S:TTG1-GR exhibited similar trichome and flavonoid phenotypes in leaves with ttg1-13 ( Figures S1 and S2). This indicated that the TTG1-GR fusion protein has a biological function like that of the wild type TTG1 upon steroid induction.
By utilizing the established steroid-inducible activation of TTG1, we further detected whether the expression of these downregulated genes is activated by TTG1 activity (Figures 1 and 2). DEX treatment of ttg1-13 35S:TTG1-GR young shoots and expanding true leaves at 20 DAG for 1 or 3 h resulted in a significant increase in the expression of TTG2, TT8, F3'H, DFR, ANS, UGT79B1, UGT75C1, 5MAT, BLT, ANL2, PAL4, DMR6, FLS3, GSTF12, MYB5, and MYB23, compared with that of the mock-treated controls ( Figure 3 and Figure S3). To determine whether 35S:TTG1-GR directly or indirectly represses these genes, we repeated DEX applications to the 35S:TTG1-GR young shoots and expanding true leaves of the ttg1-13 background in the presence of the protein synthesis inhibitor cycloheximide (CYC), because the induction of TTG1-GR activity by DEX does not require protein synthesis. We found that the combined treatment of DEX and CYC for 1 or 3 h only dramatically promoted the expression of TTG2, TT8, F3'H, DFR, ANS, UGT79B1, UGT75C1, 5MAT, and BLT ( Figure 3 and Figure S3). These indicated that these 9 genes are immediate targets transcriptionally induced by the TTG1-dependent MBW complex in the young shoots and expanding true leaves (Figure 3), whereas the activation of the other 12 genes in the ttg1-13 mutant is reliable on other intermediate proteins ( Figure S3).  Table 2 and S1). It is notable that 4 regulatory genes involved in both trichome formation and flavonoid accumulation, namely MYC2, GL2, TTG2, and TT8, were dramatically downregulated in the ttg1-13 young shoots and expanding true leaves at 18 and 22 DAG (Figures 1 and 2; Tables 1, 2, and S1). Previous studies have demonstrated that TTG2, TT8, DFR, F3'H, and UGT79B1 are targets of TTG1 [25,44,69], which is a proof of our successful gene expression analysis in this study. Taken together, these results further indicated that the TTG1-dependent MBW complex promotes trichome formation and flavonoid accumulation by activating a series of regulatory and structural genes involved in trichome development and flavonoid biosynthesis, respectively, in A. thaliana young shoots and expanding true leaves. Dynamic expression analysis of differentially expressed genes involved in trichome formation in the young shoots and expanding true leaves between the wild type ( ) and ttg1-13 ( ) plants at 18,20, and 22 days after germination. The house-keeping gene ACTIN7 was used as the internal control. The expression of each gene was first calculated relative to ACTIN7 and then normalized to its expression level at 18 days after germination in the wild type that was set to 1. Asterisks indicate significant differences in gene expression compared with the wild type control (two-tailed paired Student's t test, p ≤ 0.05). Values are means ± SD (n = 3). Error bars denote SD. Figure 2. Dynamic expression analysis of downregulated genes involved in flavonoid biosynthesis in the young shoots and expanding true leaves between the wild type and ttg1-13 plants at 18,20, and 22 days after germination. The house-keeping gene ACTIN7 was used as the internal control. The expression of each gene was first calculated relative to ACTIN7 and then normalized to its expression level at 18 days after germination in the wild type that was set to 1. Asterisks indicate significant differences in gene expression compared with the wild type control (two-tailed paired Student's t test, p ≤ 0.05). Values are means ± SD (n = 3). Error bars denote SD.

Identification of Direct Downstream Targets Contributing to Trichome Formation and Flavonoid Accumulation Regulated by the TTG1-Dependent MBW Complex in Young Shoots and Expanding True Leaves
To investigate how the TTG1-dependent MBW complex controls the mRNA expression of downstream targeted genes, we created a steroid-inducible version of TTG1 in the background of ttg1-13, in which the TTG1 gene was fused to the rat glucocorticoid receptor (GR) and driven by the 35S promoter. We isolated a ttg1-13 35S:TTG1-GR transgenic line, which fully rescued the phenotypes of ttg1-13 leaves without trichomes and flavonoids after DEX treatment every other day after germination ( Figure S2), whereas the mock-treated ttg1-13 35S:TTG1-GR exhibited similar trichome and flavonoid phenotypes in leaves with ttg1-13 ( Figures S1 and S2). This indicated that the TTG1-GR fusion protein has a biological function like that of the wild type TTG1 upon steroid induction.
By utilizing the established steroid-inducible activation of TTG1, we further detected whether the expression of these downregulated genes is activated by TTG1 activity (Figures 1 and 2 Dynamic expression analysis of differentially expressed genes involved in trichome formation in the young shoots and expanding true leaves between the wild type ( ) and ttg1-13 ( ) plants at 18,20, and 22 days after germination. The house-keeping gene ACTIN7 was used as the internal control. The expression of each gene was first calculated relative to ACTIN7 and then normalized to its expression level at 18 days after germination in the wild type that was set to 1. Asterisks indicate significant differences in gene expression compared with the wild type control (two-tailed paired Student's t test, p ≤ 0.05). Values are means ± SD (n = 3). Error bars denote SD.

Figure 2. Dynamic expression analysis of downregulated genes involved in flavonoid biosynthesis in
the young shoots and expanding true leaves between the wild type and ttg1-13 plants at 18, 20, and 22 days after germination. The house-keeping gene ACTIN7 was used as the internal control. The expression of each gene was first calculated relative to ACTIN7 and then normalized to its expression level at 18 days after germination in the wild type that was set to 1. Asterisks indicate significant differences in gene expression compared with the wild type control (two-tailed paired Student's t test, p ≤ 0.05). Values are means ± SD (n = 3). Error bars denote SD.

Identification of Direct Downstream Targets Contributing to Trichome Formation and Flavonoid Accumulation Regulated by the TTG1-Dependent MBW Complex in Young Shoots and Expanding True Leaves
To investigate how the TTG1-dependent MBW complex controls the mRNA expression of downstream targeted genes, we created a steroid-inducible version of TTG1 in the background of ttg1-13, in which the TTG1 gene was fused to the rat glucocorticoid receptor (GR) and driven by the 35S promoter. We isolated a ttg1-13 35S:TTG1-GR transgenic line, which fully rescued the phenotypes of ttg1-13 leaves without trichomes and flavonoids after DEX treatment every other day after germination ( Figure S2), whereas the mock-treated ttg1-13 35S:TTG1-GR exhibited similar trichome and flavonoid phenotypes in leaves with ttg1-13 ( Figures S1 and S2). This indicated that the TTG1-GR fusion protein has a biological function like that of the wild type TTG1 upon steroid induction.
By utilizing the established steroid-inducible activation of TTG1, we further detected whether the expression of these downregulated genes is activated by TTG1 activity (Figures 1 and 2). DEX treatment of ttg1-13 35S:TTG1-GR young shoots and expanding true leaves at 20 DAG for 1 or 3 h resulted in a significant increase in the expression of TTG2, TT8, F3'H, DFR, ANS, UGT79B1, UGT75C1, 5MAT, BLT, ANL2, PAL4, DMR6, FLS3, GSTF12, MYB5, and MYB23, compared with that of the mock-treated controls (Figures 3 and S3). To determine whether 35S:TTG1-GR directly or indirectly represses these genes, we repeated DEX applications to the 35S:TTG1-GR young shoots and expanding true leaves of the ttg1-13 background in the presence of the protein synthesis inhibitor ) plants at 18,20, and 22 days after germination. The house-keeping gene ACTIN7 was used as the internal control. The expression of each gene was first calculated relative to ACTIN7 and then normalized to its expression level at 18 days after germination in the wild type that was set to 1. Asterisks indicate significant differences in gene expression compared with the wild type control (two-tailed paired Student's t test, p ≤ 0.05). Values are means ± SD (n = 3). Error bars denote SD.  The expression of each gene was first calculated relative to ACTIN7 and then normalized to its expression level at 18 days after germination in the wild type that was set to 1. Asterisks indicate significant differences in gene expression compared with the wild type control (two-tailed paired Student's t test, p ≤ 0.05). Values are means ± SD (n = 3). Error bars denote SD.

Identification of Direct Downstream Targets Contributing to Trichome Formation and Flavonoid Accumulation Regulated by the TTG1-Dependent MBW Complex in Young Shoots and Expanding True Leaves
To investigate how the TTG1-dependent MBW complex controls the mRNA expression of downstream targeted genes, we created a steroid-inducible version of TTG1 in the background of  The expression of each gene was first calculated relative to ACTIN7 and then normalized to its expression level at 18 days after germination in the wild type that was set to 1. Asterisks indicate significant differences in gene expression compared with the wild type control (two-tailed paired Student's t test, p ≤ 0.05). Values are means ± SD (n = 3). Error bars denote SD.

Identification of Direct Downstream Targets Contributing to Trichome Formation and Flavonoid Accumulation Regulated by the TTG1-Dependent MBW Complex in Young Shoots and Expanding True Leaves
To investigate how the TTG1-dependent MBW complex controls the mRNA expression of  The expression of each gene was first calculated relative to ACTIN7 and then normalized to its expression level at 18 days after germination in the wild type that was set to 1. Asterisks indicate significant differences in gene expression compared with the wild type control (two-tailed paired Student's t test, p ≤ 0.05). Values are means ± SD (n = 3). Error bars denote SD.

Identification of Direct Downstream Targets Contributing to Trichome Formation and Flavonoid Accumulation Regulated by the TTG1-Dependent MBW Complex in Young Shoots and Expanding True Leaves
To investigate how the TTG1-dependent MBW complex controls the mRNA expression of ) ttg1-13 plants at 18,20, and 22 days after germination. The house-keeping gene ACTIN7 was used as the internal control. The expression of each gene was first calculated relative to ACTIN7 and then normalized to its expression level at 18 days after germination in the wild type that was set to 1. Asterisks indicate significant differences in gene expression compared with the wild type control (two-tailed paired Student's t test, p ≤ 0.05). Values are means ± SD (n = 3). Error bars denote SD. synthesis. We found that the combined treatment of DEX and CYC for 1 or 3 h only dramatically promoted the expression of TTG2, TT8, F3'H, DFR, ANS, UGT79B1, UGT75C1, 5MAT, and BLT (Figures 3 and S3). These indicated that these 9 genes are immediate targets transcriptionally induced by the TTG1-dependent MBW complex in the young shoots and expanding true leaves (Figure 3), whereas the activation of the other 12 genes in the ttg1-13 mutant is reliable on other intermediate proteins ( Figure S3). To explore whether the TTG1-dependent MBW complex binds directly to the promoter regions of TTG2, TT8, F3'H, DFR, ANS, UGT79B1, UGT75C1, 5MAT, and BLT to regulate their expressions, the young shoots and expanding true leaves from mock-or DEX-treated ttg1-13 35S:TTG1-GR plants with were used for the ChIP assay. To cover all the possible cis-elements, sufficient pairs of primers were designed in the promoter regions of these 9 genes (Figure 4). The ChIP results showed that TTG1-GR was associated with the promoter regions near fragments 2, 3, and 4 of TTG2, fragments 5, 7, 8, and 11 of TT8, fragments 4, 9, and 10 of F3'H, fragments 2, 3, and 4 of DFR, fragment 2 of ANS, fragments 2 and 3 of UGT79B1, fragments 1 and 2 of UGT75C1, fragments 8 and 12 of 5MAT, and fragments 2 and 3 of BLT (Figure 4). These results collectively suggested that the TTG1-dependent MBW complex binds directly to the loci of TTG2, TT8, F3'H, DFR, ANS, UGT79B1, UGT75C1, 5MAT, and BLT to promote their expression. To explore whether the TTG1-dependent MBW complex binds directly to the promoter regions of TTG2, TT8, F3'H, DFR, ANS, UGT79B1, UGT75C1, 5MAT, and BLT to regulate their expressions, the young shoots and expanding true leaves from mock-or DEX-treated ttg1-13 35S:TTG1-GR plants with were used for the ChIP assay. To cover all the possible cis-elements, sufficient pairs of primers were designed in the promoter regions of these 9 genes (Figure 4). The ChIP results showed that TTG1-GR was associated with the promoter regions near fragments 2, 3, and 4 of TTG2, fragments 5, 7, 8, and 11 of TT8, fragments 4, 9, and 10 of F3'H, fragments 2, 3, and 4 of DFR, fragment 2 of ANS, fragments 2 and 3 of UGT79B1, fragments 1 and 2 of UGT75C1, fragments 8 and 12 of 5MAT, and fragments 2 and 3 of BLT (Figure 4). These results collectively suggested that the TTG1-dependent MBW complex binds directly to the loci of TTG2, TT8, F3'H, DFR, ANS, UGT79B1, UGT75C1, 5MAT, and BLT to promote their expression.

Discussion
Transcriptional regulation is considered to be the essential mechanism controlling the gene expression of metabolic pathways in higher plants. Extensive studies have elucidated the TTG1dependent MBW complex that controls trichome formation and flavonoid accumulation in A. thaliana. However, its direct targeted genes are still largely unknown in the regulation of trichome formation and flavonoid accumulation in A. thaliana. In this study, we identified a series of new downstream targeted genes for trichome formation and flavonoid accumulation directly or indirectly regulated by the TTG1-dependent MBW complex at the genome-wide level in A. thaliana young shoots and expanding true leaves.
We found that there were 4 downregulated regulatory genes involved in both trichome formation and flavonoid biosynthesis, TTG2, TT8, GL2, and MYC2 (Figures 1 and 2; Tables 1 and 2), among which the former two genes and the latter two genes were directly and indirectly regulated by the TTG1-dependent MBW complex (Figures 3 and 4, and Figure S3), in the ttg1-13 young shoots and expanding true leaves. The TTG2 protein encoding a WRKY transcription factor plays positive roles in trichome initiation and flavonoid accumulation and is also directly induced by TTG1 [25,57], which is consistent with our results (Figures 3 and 4). The TT8 transcription factor is essential for trichome initiation in the margin of true leaves [62] and positively regulates flavonoid biosynthesis [68,69]. GL2 functions as a transcriptional activator and repressor in trichome initiation [22][23][24] and anthocyanin biosynthesis [66], respectively. The previous study showed that GL2 inhibits anthocyanin biosynthesis by directly repressing the expression of PAP1, PAP2, MYB113, MYB114, and TT8 [66], which theoretically should be increased because of the lower expression of GL2 in the ttg1-13 young shoots and expanding true leaves. However, the TT8 expression was significantly reduced (Figures 1 and 2; Tables 1 and 2,  and Table S1), and the expression of PAP1, PAP2, MYB113, and MYB114 was not altered (Tables S1 and  S2). Therefore, it can be deduced that the TTG1-dependent MBW complex promotes TT8 expression probably independent of GL2, or the simulative effect of the TTG1-dependent MBW complex and/or other downregulated regulatory genes is stronger than the inhibitory effect of GL2 on the TT8 expression, in the young shoots and expanding true leaves. Similarly, the simulative effect of the TTG1-dependent MBW complex and/or other downregulated regulatory genes is comparable to the inhibitory effect of GL2 on the expression of PAP1, PAP2, MYB113, and MYB114. These interesting questions need further investigation. Genetic analysis showed that the double mutant of ttg2-1 gl2-1 has more defective trichome development than in either single mutant, and TTG2 positively regulates GL2 [57]. Thus, it is possible that the reduced expression of GL2 is caused by the lower expression of TTG2, which would be in accordance with GL2 being indirectly regulated by the TTG1-dependent MBW complex ( Figure S3). The bHLH transcription factor MYC2 as the central mediator of the phytohormone jasmonic acid (JA) signaling positively regulates trichome initiation and flavonoid accumulation [50][51][52][53]. The previous study indicated that all the three phytohormones inclusive of JA, gibberellin A3 (GA3), and cytokinin (6-benzylaminopurine) stimulate trichome initiation not through regulating the TTG1 expression, and whereas both JA and 6-benzylaminopurine promote anthocyanin production, GA3 does not [62]. The expression of JA biosynthetic genes was not altered in the ttg1-13 young shoots and expanding true leaves (Tables S1 and S2), indicating that the TTG1-dependent MBW complex and JA function in an independent manner in inducing the expression of MYC2, thus promoting trichome formation and flavonoid accumulation. Gibberellin 2-β-DIOXYGENASE6 (GA2OX6) encoding a GA 2-oxidase functions in converting bioactive gibberellins and their precursors into inactive forms, thus reducing endogenous bioactive gibberellins [83]. CYTOKININ DEHYDROGENASE4 (CKX4) encodes a cytokinin oxidase that is responsible for the cytokinin accumulation [84]. The expression levels of GA2OX6 and CKX4 were significantly decreased in the ttg1-13 young shoots and expanding true leaves (Table S1), suggesting that the TTG1-dependent MBW complex promotes trichome formation, and flavonoid accumulation might rely on cytokinin, but not gibberellins.
We demonstrated that the TTG1-dependent MBW complex directly promoted the expression of the regulatory gene BLT (Figures 3 and 4) and indirectly activated the expression of the regulatory genes SPL8, ETC1, HDG11, CPC, MYB106, MYB5, SIM, and MYB23, and the structural gene SVB during trichome formation in the young shoots and expanding true leaves ( Figure S3). BLT encodes a key regulator of trichome branching, and its mutation results in the formation of branchless trichomes with blunt tips [54,55]. SPL8 positively regulates the trichome number on flower sepals [48]. ETC1 acts in concert with CPC to repress the trichome cell fate in the shoot epidermis [49]. SVB is positively correlated with trichome size, and its mutant exhibits trichome branches of variable length and number [65]. HDG11 encodes a homeodomain leucine zipper transcription factor like GL2 and displays a positive role in trichome differentiation [56]. MYB106 functions as an activator in trichome differentiation [58], and a repressor in trichome branching [59]. Loss of functions of MYB5 and MYB23 display increased numbers of small and two-branched trichomes, respectively [60,61]. Further genetic analysis indicated that MYB5 and MYB23 are partially redundant in repressing trichome branching [60]. MYB23 is directly activated by GL2 during trichome formation [56], suggesting that the TTG1-dependent MBW complex promotes the expression of MYB23 possibly via the direct activation of GL2. SIM encodes a member of plant-specific cyclin-dependent kinase inhibitors, and its mutation results in multicellular trichomes [63,64]. Considering that there is no trichome present in the ttg1-13 young shoots and expanding true leaves ( Figure S1), it is plausible to assume that the TTG1-dependent MBW complex promotes trichome formation first by activating the expression of the genes that induce trichome initiation and differentiation, including TTG2, GL2, MYC2, SPL8, HDG11, and MYB106, and then by promoting the other genes that positively regulates branching, size, and cell number, including BLT, MYB5, SIM, MYB23, and SVB, in the young shoots and expanding true leaves. Further, the TTG1-dependent MBW complex might be independent of ETC1 and CPC or might be due to the fact that the negative influence of the two genes is noncompetitive with the positive effect of other genes during trichome formation in the young shoots and expanding true leaves.
On the other hand, we proved that the TTG1-dependent MBW complex directly promoted the expression of the regulatory gene TT8 and the structural genes F3'H, DFR, ANS, UGT79B1, UGT75C1, and 5MAT (Figures 3 and 4), and indirectly activated the expression of the regulatory gene ANL2 and the structural genes PAL4, DMR6, FLS3, and GSTF12 during flavonoid biosynthesis in the young shoots and expanding true leaves ( Figure S3). PAL comprises four isoforms of PAL1, PAL2, PAL3, and PAL4 that convert phenylalanine into trans-cinnamic acid, which is the first step in the flavonoid biosynthetic pathway in A. thaliana [70,71]. F3'H encoding a cytochrome P450 monooxygenase catalyzes the formation of eriodictyol and dihydroquercetin from naringenin and dihydrokaempferol, respectively [70,72]. DMR6 is a flavone synthase I enzyme that catalyzes the conversion of the flavanones into flavones [73]. FLS acts as the first committed enzyme for flavonol biosynthesis [85,86], and there are five FLS genes in the A. thaliana genome, among which FLS3 exhibits the FLS activity, promoting flavonol accumulation [74]. DFR catalyzes the formation of leucoanthocyanidins from dihydroflavonols [2,75]. ANS encoding a 2-oxoglutarate-dependent dioxygenase catalyzes the conversion of leucoanthocyanidins into 3-OH-anthocyanins [76,77]. UGT79B1 is involved in the glycosylation of anthocyanins at the C-5 position, and its knockout mutant contains a drastically decreased content of anthocyanins [70,78]. Another UDP-glucose: UGT75C1 is involved in the malonylation of anthocyanins, and its mutation results in the complete loss of anthocyanin 5-O-glucosides [79]. 5MAT is specific for malonyl-CoA and for anthocyanins with 5-O-glucosylation and accelerates the accumulation of malonylated anthocyanins [80]. GSTF12, a member of GST-like proteins, is involved in transport and promotes anthocyanin accumulation [81,82]. ANL2, as a member of homeodomain proteins, positively regulates anthocyanin accumulation in shoot cells [67]. These suggested that the TTG1-dependent MBW complex promotes flavonoid biosynthesis directly or indirectly through a series of regulatory and structural genes, and TTG1-mediated flavonoid biosynthesis might be independent of GL2 or might be due to the fact that the negative influence of GL2 is noncompetitive with the positive effect of other genes, in the young shoots and expanding true leaves.
Previous studies indicated that the TTG1-dependent MBW complex has no obvious effect on MYB5 expression in seeds [60] and directly activates BAN expression in siliques [46]. Here, we found that MYB5 and BAN were indirectly activated and not regulated, respectively, in the ttg1-13 young shoots and expanding true leaves ( Figure S3; Tables S1 and S2), indicating that the TTG1-dependent MBW complex regulates gene expression is tissue-dependent during trichome formation and flavonoid accumulation. It is worth noting that 8 and 7 downregulated genes involved in flavonoid accumulation were directly and indirectly regulated by the TTG1-dependent MBW complex, whereas 2 and 11 downregulated genes related to trichome formation were directly and indirectly regulated by the TTG1-dependent MBW complex in the young shoots and expanding true leaves (Figures 3 and 4, and Figure S3; Tables 1 and 2, Tables S1 and S2), suggesting that the TTG1-dependent MBW complex is likely to be a key direct and indirect transcription factor in regulating flavonoid accumulation and trichome formation, respectively.
In the present study, we report on direct targets of the TTG1-dependent MBW complex, revealing that the TTG1-dependent MBW complex functions primarily or exclusively as a transcriptional direct and indirect activator in the regulation of trichome formation and flavonoid accumulation, respectively, in the young shoots and expanding true leaves of A. thaliana.

Plant Material and Growth Condition
The A. thaliana ecotype Columbia (Col-0) was used as the wild type control. The ttg1-13 mutant was utilized in our previous study [14]. The growth condition of A. thaliana plants has been described previously [14].

Generation of Transgenic Plants
The construct of 35S:TTG1-GR, which was created in our previous study [7], was transformed into the ttg1-13 mutant via floral dip [87]. The ttg1-13 35S:TTG1-GR transgenic plants were selected by Basta on soil and also verified by DNA analysis until T3 homozygous transgenic progeny was generated.

RNA Sequencing (RNA-seq) and Data Analyses
The samples of young shoots and expanding true leaves used for RNA-seq analysis were carefully harvested from Col-0 and ttg1-13 plants at 20 DAG following the previously reported method [88]. Three independent biological replicates from three different plantings were performed in the RNA-seq experiment. The RNA-seq and data analysis were carried out through the Gene Denovo service (http://www.genedenovo.com/) following the standard protocol (http://www.genedenovo.com/product/ 41.html): (1) The quality and quantity of the isolated RNA samples were assessed by Nanodrop 2000 Spectrophotometer (Thermo, Wilmington, NC, USA) and Agilent 2100 Bioanalyzer (Agilent, Böblingen, Germany), (2) removement of the possible DNA contamination with DNase I, (3) mRNA enrichment and fragmentation, (4) sequencing adaptor ligation and PCR amplification, and (5) assessment on quality and quantity of the sample library. Finally, the cDNA library products were utilized for sequenced analysis via the Illumina HiSeq™ 2000. Transcript abundance was calculated as RPKM (reads per Kb per million reads) [89]. RPKM values presenting as "0" were artificially set to "0.001" for subsequent analysis. Comparisons of RPKM between treatments (WT (Col-0) vs. ttg1-13) were performed for each Unigene. The DEGs were functionally classified using the biological process category of Arabidopsis Gene Ontology (http://www.geneontology.org). A differential expression analysis of the two treatments was conducted using the DESeq R package (1.10.1). The resulting p-values were adjusted using Benjamini and Hochberg's approach for controlling the false discovery rate (FDR). The DEGs with |log 2 ratios| ≥ 1 and FDR ≤ 0.05 are functionally categorized and listed in Tables S1 and S2.

Gene Expression Analysis
The tissues of young shoots and expanding true leaves used for gene expression analysis are the same as those in the RNA-seq experiment. They were collected from eight individual plants at 20 DAG, and three independent biological replicates were performed for gene expression analysis. Total RNA extraction and first-strand cDNA synthesis were conducted according to the method previously described [90]. The SYBR Green Master Mix (TaKaRa Bio, Dalian, China) was utilized for qRT-PCR reaction with QuantStudio TM 7 Flex Real-Time PCR System (Life technologies, Carlsbad, CA, USA), and the relative gene expression level was calculated as reported previously [91]. The house-keeping gene ACTIN7 was used as the internal control. Primers used for qRT-PCR analysis are listed in Table S3.

GR Induction
For the induction of TTG1:GR, young shoots and expanding true leaves were harvested from 8 individual plants at 20 DAG 1 and 3 h after different treatments of mock, DEX, CYC, and DEX plus CYC. The concentrations of DEX and CYC used here were 10 and 5 µM, respectively.

ChIP Assay
The tissues of young shoots and expanding true leaves at 20 DAG used for ChIP assay are the same as those in the RNA-seq experiment. The ttg1-13 35S:TTG1-GR plants after germination were treated for 10 days with mock and 10 µM DEX every other day, after which they were used for the ChIP experiment. The ChIP assay was carried out as described previously [7]. In brief, 3 g of young shoots and expanding true leaves were fixed with 37% formaldehyde. After nuclear protein-DNA extraction and sonication, immunoprecipitation was conducted using a GR antibody coupled to magnetic beads. The relative enrichment of each fragment was detected by qRT-PCR, and the ChIP experiment was performed in three biological replicates. Primers used for ChIP assay are listed in Table S3.

Statistical Analysis
Completely randomized block designs were utilized in three biological replicates. Data were analyzed with use of the SPSS statistical package (version 8.0). The two-tailed paired Student's t-test was used to analyze gene expression. p values ≤ 0.05 indicated a statistically significant difference.
Supplementary Materials: Supplementary materials can be found at http://www.mdpi.com/1422-0067/20/20/ 5014/s1. Supplementary Figure S1. The schematic diagram shows tissues from the wild type (Col-0) and ttg1-13 plants at 20 days after germination harvested for the RNA-seq experiment. The yellow crosses indicate that the whole leaves, including blades and petioles, were removed completely, whereas the red ellipses indicate that the tissues, including young shoots and expanding true leaves, were collected for RNA-seq analysis. Supplementary Figure S2. Generation of a ttg1-13 35S:TTG1-GR transgenic line containing the biologically active TTG1-GR fusion, and the tissues of young shoots and expanding true leaves indicated by the red ellipses were collected for GR induction and ChIP analyses. Individual seedlings of ttg1-13 35S:TTG1-GR plants were mock-treated (Mock, left) or treated with 10 µM dexamethasone (Dex, right) every other day after germination. Mock plants generate leaves without flavonoids and trichomes as ttg1-13, whereas DEX treatment rescues these phenotypes. Supplementary Figure S3. The genes involved in flavonoid biosynthesis and trichome formation are not immediate targets of transcriptional regulation by TTG1 in the young shoots and expanding true leaves. The ttg1-13 35S:TTG1-GR young shoots and expanding true leaves at 20 days after germination were mock-treated (Mock) or treated with 10 µM dexamethasone (DEX), 5 µM cycloheximide (CYC), or 10 µM DEX plus 5 µM CYC (DEX + CYC). The expression of these genes was determined after 1 or 3 h of treatment using qRT-PCR analyses. The house-keeping gene ACTIN7 was used as the internal control. The expression level of each gene was first calculated relative to ACTIN7, and its expression levels in both Mock and CYC treatments were set to 1. Asterisks indicate significant differences in gene expression in DEX-treated samples compared with their respective controls (two-tailed paired Student's t test, p ≤ 0.05). Values are means ± SD (n = 3). Error bars denote SD. Supplementary Table S1. A list of genes expressed at lower levels ('downregulated') in the young shoots and expanding true leaves of ttg1-13 plants than in wild type rosette leaves at 20 days after germination. Supplementary Table S2. A list of genes expressed at higher levels ('upregulated') in the young shoots and expanding true leaves of ttg1-13 plants than in wild type rosette leaves at 20 days after germination. Supplementary Table S3. Primers used in the present study.

Conflicts of Interest:
The authors declare no conflict of interest. Besides, the funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.