Transcription factors BZR1 and PAP1 cooperate to promote anthocyanin biosynthesis in Arabidopsis shoots

Abstract

Anthocyanins play critical roles in protecting plant tissues against diverse stresses. The complicated regulatory networks induced by various environmental factors modulate the homeostatic level of anthocyanins. Here, we show that anthocyanin accumulation is induced by brassinosteroids (BRs) in Arabidopsis (Arabidopsis thaliana) shoots and shed light on the underlying regulatory mechanism. We observed that anthocyanin levels are altered considerably in BR-related mutants, and BRs induce anthocyanin accumulation by upregulating the expression of anthocyanin biosynthetic genes. Our genetic analysis indicated that BRASSINAZOLE RESISTANT 1 (BZR1) and PRODUCTION OF ANTHOCYANIN PIGMENT 1 (PAP1) are essential for BR-induced anthocyanin accumulation. The BR-responsive transcription factor BZR1 directly binds to the PAP1 promoter, regulating its expression. In addition, we found that intense anthocyanin accumulation caused by the pap1-D-dominant mutation is significantly reduced in BR mutants, implying that BR activity is required for PAP1 function after PAP1 transcription. Moreover, we demonstrated that BZR1 physically interacts with PAP1 to cooperatively regulate the expression of PAP1-target genes, such as TRANSPARENT TESTA 8, DIHYDROFLAVONOL 4-REDUCTASE, and LEUKOANTHOCYANIDIN DIOXYGENASE. Our findings indicate that BZR1 functions as an integral component of the PAP1-containing transcription factor complex, contributing to increased anthocyanin biosynthesis. Notably, we also show that functional interaction of BZR1 with PAP1 is required for anthocyanin accumulation induced by low nitrogen stress. Taken together, our results demonstrate that BR-regulated BZR1 promotes anthocyanin biosynthesis through cooperative interaction with PAP1 of the MBW complex.

A brassinosteroid-responsive transcription factor induces the expression of anthocyanin biosynthetic genes thorough the functional interaction with the Production of Anthocyanin Pigment 1.

Introduction

Anthocyanins are a class of water-soluble flavonoids bearing sugar groups. These pigments provide red, purple, and blue colors to plant organs, which attracts pollinators. In addition, anthocyanins play essential roles in plant tissue protection against environmental stresses, such as high light, UV irradiation, cold temperature, and nutrient starvation (Holton and Cornish 1995; Steyn et al. 2002; Ramakrishna and Ravishankar 2011; Naing and Kim 2021). Dietary anthocyanins provide anti-inflammatory benefits, boosting the immune system in humans (Meiers et al. 2001; Zhang et al. 2014).

In plants, the flavonoid pathway for anthocyanin biosynthesis is initiated by CHALCONE SYNTHASE (CHS), which mediates the condensation of 3 molecules of malonyl CoA and one of p-coumaroyl CoA derived from a general phenylpropanoid pathway. In Arabidopsis, the anthocyanin biosynthetic pathway is divided into early and late biosyntheses. The early biosynthetic genes (EBGs) include CHS, CHALCONE ISOMERASE (CHI), and FLAVANONE 3-HYDROXYLASE (F3H), whereas the late biosynthetic genes (LBGs) are FLAVONOID 3′-HYDROXYLASE (F3′H), DIHYDROFLAVONOL 4-REDUCTASE (DFR), LEUKOANTHOCYANIDIN DIOXYGENASE (LDOX), and UDP-GLUCOSE FLAVONOID 3-O-GLUCOSYLTRANSFERASE (UF3GT) (Dooner et al. 1991; Holton and Cornish 1995). The EBGs are involved in common flavonoid biosynthesis and are regulated transcriptionally by 3 redundant R2R3-MYB transcription factors MYB11, MYB12, and MYB111. In contrast, the LBGs practically contribute to producing anthocyanins and their transcript levels are regulated primarily by the MBW protein complexes composed of 3 transcription factors: MYB, bHLH, and WD40-REPEAT PROTEIN (WDR) (Shirley et al. 1995; Pelletier et al. 1997; Gonzalez et al. 2008).

In addition to anthocyanin biosynthesis, the MBW protein complex in Arabidopsis regulates the production of seed-coat mucilage, proanthocyanidin accumulation, and development of root-hair and trichome (Payne et al. 2000; Zhang et al. 2003; Gonzalez et al. 2009; Xu et al. 2014; Jiang et al. 2024). While specific MYB proteins have distinct functions in different MBW-mediated pathways, the bHLH transcription factors, including TRANSPARENT TESTA 8 (TT8), GLABRA 3 (GL3), and ENHANCER OF GLABRA 3 (EGL3), and the WDR protein, TRANSPARENT TESTA GLABRA 1 (TTG1), function redundantly. The MYB transcription factors that positively regulate anthocyanin production include PRODUCTION OF ANTHOCYANIN PIGMENT 1 (PAP1)/MYB75, PAP2/MYB90, MYB113, and MYB114 (Zhang et al. 2003; Broun 2005; Zhang et al. 2019). Of these, transcriptional regulation of anthocyanin biosynthesis by the MBW complex driven by PAP1 is sufficiently well-known. In contrast, several MYB proteins act as repressors for anthocyanin biosynthesis. For example, MYBL2 binds to bHLH, repressing transcriptional activity of the MBW complex (Matsui et al. 2008).

Anthocyanin contents in plant tissues are regulated primarily by various abiotic stresses such as high light, UV, cold, drought, excess sucrose supply, and nutrient depletion. LONG HYPOCOTYL 5 (HY5) directly binds to the PAP1 promoter and promotes PAP1 expression under far-red and blue light. HY5 also directly induces the expression of LBGs such as DFR and LDOX in high light conditions (Shin et al. 2013). In temperature-regulated anthocyanin production, a CONSTITUTIVE PHOTOMORPHOGENIC 1 (COP1)-HY5 module plays a critical role in the transcriptional regulation of anthocyanin biosynthetic genes (Kim et al. 2017). In addition, dynamic changes in anthocyanin contents by changes in light conditions are mediated by posttranslational regulation of PAP1. At the protein level, PAP1 is stabilized under light by MAP kinase-mediated phosphorylation or SAP AND MIZ1 DOMAIN-CONTAINING LIGASE 1 (SIZ1)-induced sumoylation, whereas PAP1 is degraded by the E3 Ubiquitin ligase COP1 in the dark (Maier et al. 2013; Li et al. 2016; Zheng et al. 2020). High concentrations of sucrose promote anthocyanin accumulation by upregulating the expression of genes such as GL3, TT8, and PAP1 (Solfanelli et al. 2005; Jeong et al. 2010). Three nitrogen-induced LATERAL ORGAN BOUNDARY DOMAIN (LBD) family transcription factors (LBD37, LBD38, and LBD39) suppress the expression of PAP1 and PAP2 in the presence of a nitrogen source. Nitrogen depletion represses the expression of the LBD genes, inducing anthocyanin accumulation mediated by PAP1 and PAP2 (Rubin et al. 2009).

The intrinsic hormonal signals also contribute to regulate anthocyanin biosynthesis. The JASMONATE-ZIM DOMAIN (JAZ) protein, acting as a repressor of jasmonic acid (JA) signaling, directly binds to PAP1 and 3 bHLH proteins, inhibiting the formation of the MBW complex (Qi et al. 2011). Strigolactone promotes anthocyanin biosynthesis through inhibition of SUPPRESSOR OF MAX2 1-LIKE 6 (SMXL6) acting as a direct transcriptional repressor of the PAP1 gene (Wang et al. 2020). In contrast, DELLA repressors of gibberellin (GA) signaling positively regulate anthocyanin production via sequestering MYBL2 and JAZ repressors (Xie et al. 2016). While cytokinin increases the transcript level of PAP1 and decreases MYBL2 expression in high sucrose conditions, ethylene suppresses the sucrose-induced anthocyanin accumulation by downregulating the expression of the sucrose transporter SUC1 (Jeong et al. 2010; Das et al. 2012).

Brassinosteroids (BRs) are a class of plant steroid hormones that play pivotal roles in diverse physiological and developmental processes such as cell growth, photomorphogenesis, stomatal movement and development, vascular differentiation, and reproductive organ development (Caño-Delgado et al. 2004; Ye et al. 2010; Clouse 2011; Kim et al. 2012; Ha et al. 2016). The biological activities of BRs are regulated by gene expression mediated by the BR-responsive transcription factors including BRASSINAZOLE RESISTANT 1 (BZR1) and BRI1-EMS-SUPPRESSOR 1 (BES1) (Wang et al. 2002; Yin et al. 2002). Genome-wide analysis showed that BZR1 and BES1 tend to bind the E-box (CANNTG) motif and BR-Response Element (CGTG^T/_CG) on the promoters of BR-upregulated and BR-downregulated genes, respectively (Sun et al. 2010; Yu et al. 2011). In addition, BZR1 and BES1 also cooperate with other transcription factors in the crosstalk with light, auxin, and GA signaling (Nolan et al. 2020).

Previous studies have shown that BR can also modulate anthocyanin contents. The JA- or cytokinin-induced anthocyanin accumulation is hindered in BR-deficient conditions, whereas exogenous BR strengthens the effect of JA and cytokinin on anthocyanin contents in Arabidopsis (Peng et al. 2011; Yuan et al. 2015). In grape vines (Vitis vinifera), BR application promotes fruit ripening and anthocyanin production in grape skin (Symons et al. 2006). However, the fundamental role of BR in anthocyanin production is still unclear and the molecular mechanism underlying BR-mediated anthocyanin biosynthesis remains unknown.

In this study, we demonstrate that BR promotes anthocyanin accumulation through the dual regulation of PAP1 by BZR1 in Arabidopsis seedlings. The BZR1 transcription factor not only induces the expression of the PAP1 gene but also directly interacts with PAP1 proteins, promoting the expression of anthocyanin biosynthetic genes. Our genetic and biochemical analyses indicate that the functional interaction of BZR1 with PAP1 cooperatively regulates anthocyanin biosynthesis and BZR1 enhances the transcriptional activity of the MBW protein complex. Furthermore, we found that PAP1 regulation by BZR1 is required for anthocyanin accumulation induced by low nitrogen conditions.

Results

BRs promote anthocyanin accumulation in Arabidopsis seedlings

When Arabidopsis seedlings were grown on brassinolide (BL, the most active BR)-containing medium, accumulation of anthocyanin in shoots was often observed. This led us to carefully examine whether BR modulates anthocyanin production in vegetative tissues of plants. We first measured anthocyanin contents in the BR-deficient mutants, cyp85a1a2 and Sdet2 (Park et al. 2014). In 3-d-old seedlings grown on half-strength Murashige and Skoog (½ MS) medium containing 1 μM norflurazon, the level of anthocyanin accumulation at the junction between the hypocotyls and cotyledons was lower in both cyp85a1a2 and Sdet2 mutants compared to the wild type (Fig. 1A). In shoots of 14-d-old seedlings, the anthocyanin level of cyp85a1a2 and Sdet2 was much lower than that of wild-type Col-0 (Fig. 1B). We additionally noted anthocyanin accumulation in various regions, including the inflorescence, and rosette-stem junction, during the adult stage of plant growth. Anthocyanin accumulation was reduced at the stem-rosette junction, but not inflorescence, of cyp85a1a2 and Sdet2 mutants compared to the wild type (Supplementary Fig. S1).

Figure 1.

BR promotes anthocyanin production in Arabidopsis seedlings. A) Anthocyanin accumulation in early seedlings of wild-type Col-0, Sdet2, and cyp85a1a2. Seedlings were grown on ½ MS medium containing 1 µM norflurazon for 3 d. Black square brackets indicate anthocyanin-accumulated region in hypocotyl. Scale bars indicate 10 µm. B) Comparison of anthocyanin levels in the shoots of 14-d-old wild-type Col-0, Sdet2, and cyp85a1a2 seedlings (n = 15). C) Changes in anthocyanin levels upon PCZ treatment. Anthocyanin contents were measured in the shoots of 14-d-old Col-0 grown on ½ MS medium with a series concentration of PCZ (n = 15). D) to F) Comparison of anthocyanin levels in 14-d-old Col-0, bri1-116, bri1-116 bzr1-1DD), WS, bri1-5, and bin2-3 bil1 bil2E), and Col-0, bin2-1 heterozygous (+/−), bin2-1 homozygous (−/−), and bin2-1 bzr1-1DF) (n = 15). Error bars indicate standard errors. The statistically significant differences are indicated by different lowercase letters (one-way ANOVA, P < 0.05). FW, fresh weight.

Similarly, treatment with propiconazole (PCZ), a BR biosynthesis inhibitor, reduced anthocyanin content in a dose-dependent manner (Fig. 1C). In addition, the anthocyanin contents in various mutants related to BR signaling components were measured. Two loss-of-function mutants of the BR receptor BRASSINOSTEROID INSENSITIVE 1 (BRI1), bri1-116 and bri1-5, showed reduced anthocyanin levels compared with the wild type (Fig. 1, D and E). Notably, anthocyanin levels of the bri1-116 mutant were restored by the bzr1-1D mutation (Fig. 1D). In contrast, bin2-3 bil1 bil2, the loss-of-function mutant of BIN2 and its 2 closest homologs (Yan et al. 2009), produced more anthocyanin than its wild-type WS (Fig. 1E). A comparative analysis of the anthocyanin levels in Col-0, bin2-1, and bin2-1 bzr1-1D revealed that the bzr1-1D mutation restored the reduced anthocyanin levels of the bin2-1 mutant (Fig. 1F), indicating that BR-responsive transcription factor BZR1 positively regulates anthocyanin production.

BZR1 mediates BR-regulated anthocyanin biosynthesis

Next, we found that anthocyanin accumulation in Col-0 was increased by BL (Fig. 2, A and B). Moreover, BL-induced anthocyanin accumulation was enhanced further in the bzr1-1D gain-of-function mutant. Similarly, bikinin, an inhibitor of Arabidopsis GSK3-like kinases, induced anthocyanin accumulation more strongly in bzr1-1D (Supplementary Fig. S2A). When we compared anthocyanin accumulation in 3-, 7-, 10-, and 14-d-old seedlings upon BL treatment, anthocyanin accumulation induced by BL was observed more clearly as seedlings grew (Supplementary Fig. S2B). Notably, anthocyanin levels in the bzr1-1D mutant did not show an increase under mock conditions (Fig. 2B) although the bzr1-1D mutation restored anthocyanin levels of bri1-116 and bin2-1 (Fig. 1, D and F). This is reminiscent of the growth regulation phenotype caused by bzr1-1D. Despite BZR1's role in promoting growth in response to BR, the bzr1-1D mutant does not exhibit any noticeable growth phenotype compared to the wild type (Wang et al. 2002). Nevertheless, the bzr1-1D mutation effectively rescues the severe dwarf phenotype observed in bri1-116 or bin2-1 mutants.

Figure 2.

BZR1 mediates BR-regulated anthocyanin biosynthesis. A) Phenotypes of Col-0 and bzr1-1D seedling grown on ½ MS medium without or with 50 nm BL for 14 d. Scale bars indicate 1 cm. B) Anthocyanin contents in Col-0 and bzr1-1D seedlings grown on ½ MS medium containing 0, 5, or 50 nm BL for 14 d (n = 15). FW, fresh weight. C) and D) RT-qPCR analyses for the expression level of anthocyanin biosynthetic genes C) and transcription factors composing the MBW complex D) in the shoots of Col-0 and bzr1-1D seedlings. Plants were grown on ½ MS medium without or with 50 nm BL for 14 d. The data shown are representative of 3 independent experiments. E) Phenotypes of Col-0 and the bzr-q (bzr1 bes1 beh2 beh3) mutant grown on ½ MS medium without or with 50 nm BL for 14 d. Scale bars indicate 1 cm. F) Anthocyanin contents in the shoots of Col-0 and bzr-q (n = 15). Plants were grown on ½ MS medium containing 0, 5, or 50 nm BL for 14 d. FW, fresh weight. G) Anthocyanin contents in the shoots of Col-0 and bzr-q (n = 15). Plants were grown on ½ MS medium containing 0, 5, or 10 μM bikinin. FW, fresh weight. H) RT-qPCR analysis for the expression of anthocyanin biosynthetic genes in the shoots of Col-0 and bzr-q. Plants were grown on ½ MS medium for 14 d. The data shown are representative of 3 independent experiments. All RT-qPCR data were normalized to the expression of the PP2A gene. The statistically significant differences are indicated by different lowercase letters (one-way ANOVA, P < 0.05). In RT-qPCR, ANOVA analysis was performed independently for each gene. Error bars in RT-qPCR analysis represent standard deviations and error bars in other results indicate standard errors.

Analysis of the expression of genes involved in anthocyanin biosynthesis in BL-treated Col-0 and bzr1-1D revealed that BL significantly promoted the expression of anthocyanin biosynthetic genes (CHS, CHI, F3H, F3′H, DFR, LDOX, and UF3GT) (Fig. 2C). Notably, BL also increased the expression of PAP1, PAP2, and TT8, key transcription factors upregulating the expression of anthocyanin biosynthetic genes (Fig. 2D). Consistent with the phenotype of bzr1-1D, which accumulates more anthocyanin under BL treatment than the wild-type, BL-induced gene expression was enhanced further by the bzr1-1D mutation (Fig. 2, C and D). In addition, the bes1-D gain-of-function mutant exhibited a similar pattern to bzr1-1D in anthocyanin accumulation and expression of anthocyanin biosynthetic genes (Supplementary Fig. S3).

Considering that BL increases the expression of both EBGs and LBGs, we explored the potential impact of BR on flavonol and proanthocyanidin contents. We found that cellular flavonol levels, as detected by diphenylboric acid 2-aminoethyl ester (DPBA), were increased by BL treatment (Supplementary Fig. S4, A and B). However, compared to BL-induced anthocyanin accumulation, the alterations in flavonol contents caused by BL were relatively minor. In addition, proanthocyanidin levels observed by 4-dimethylaminocinnamaldehyde (DMACA) staining showed no difference between Col-0 and BR mutants (Supplementary Fig. S4C).

To understand whether BZR1-mediated transcriptional regulation is essential for the anthocyanin production, we analyzed anthocyanin contents in bzr1 and bes1 mutant. However, compared with wild-type Col-0, both bzr1 and bes1 did not show a significant difference in anthocyanin levels, and anthocyanin accumulation induced by BL in bzr1 and bes1 was either similar to that in Col-0 (Supplementary Fig. S5A). A previous study has shown that 6 BZR1/BES1 family members function redundantly in BR signaling, and accordingly, only the bzr hextuple mutant displays severe growth defects (Chen et al. 2019). Due to the extreme dwarfism exhibited by the mutant, which hinders accurate measurements of anthocyanin content and fresh weight, we generated the bzr quadruple mutant (bzr-q; bzr1 bes1 beh2 beh3) as a moderate BZR1/BES1 loss-of-function line (Supplementary Fig. S5B).

In the measurement of anthocyanin contents, bzr-q showed a significantly reduced level of anthocyanin compared with Col-0 and little response to BL or bikinin (Fig. 2, E to G). Accordingly, BR-induced expression of anthocyanin biosynthetic genes was impaired in the bzr-q mutant (Fig. 2H). When we examined anthocyanin contents in overexpression plants for BZR1/BES1 homologs (BEH1 to BEH4), BL-induced anthocyanin accumulation was enhanced further in Pro35S:BEH1-YFP and Pro35S:BEH4-YFP (Supplementary Fig. S6A). In addition, anthocyanin accumulation in rosette leaves and the rosette-stem junction was higher in plants overexpressing BZR1/BES1 family members compared to the wild type. However, changes in anthocyanin accumulation in inflorescence of these plants were not observed (Supplementary Fig. S6B). Our results indicated that BZR1/BES1 members mediate BR-induced anthocyanin accumulation.

BR-induced anthocyanin accumulation is dependent on the activity of the MBW complex

To investigate whether BR-induced anthocyanin accumulation is connected to the activity of the MBW complex, we tested the effect of BL on the MBW complex-related mutants. It has been shown that the pap1-D mutation by activation tagging of PAP1 drastically elevates anthocyanin levels (Borevitz et al. 2000). In pap1-D, in addition to anthocyanin biosynthetic genes, 2 transcription factors, PAP2 and TT8, were considerably upregulated by the overexpression of PAP1 (Fig. 3, A and B). These gene expression patterns of pap1-D were very similar to BR-regulated gene expression (Fig. 2, C and D). Moreover, anthocyanin accumulation in the pap1-D mutants was further enhanced by BL treatment (Fig. 3, C and D). Importantly, the pap1 loss-of-function mutant was insensitive to BL and bikinin in anthocyanin accumulation (Fig. 3E), indicating that PAP1 is essential for BR-induced anthocyanin accumulation.

Figure 3.

PAP1 is indispensable for BR-induced anthocyanin accumulation. A) and B) RT-qPCR analyses for the expression level of anthocyanin biosynthetic genes A) and transcription factors composing the MBW complex B) in the shoots of 14-d-old Col-0 and pap1-D seedlings. All data were normalized to the expression of the PP2A gene. The data shown are representative of 3 independent experiments. Significant differences were determined by Student's t-test (n.s. no significant difference, and *P < 0.05). C) Phenotypes of Col-0 and pap1-D seedling grown on ½ MS medium without or with 50 nm BL for 14 d. Scale bars indicate 1 cm. D) Anthocyanin contents in the shoots of Col-0 and pap1-D seedlings grown on ½ MS medium containing 0, 5, or 50 nm BL for 14 d (n = 15). Numbers indicate the fold change value. FW, fresh weight. E) Anthocyanin contents in the shoots of wild-type No-0 and the pap1 grown on ½ MS medium without or with 50 nm BL or 10 μM bikinin for 14 d (n = 15). FW, fresh weight. Error bars in RT-qPCR results represent standard deviations and error bars in other results indicate standard errors. The statistically significant differences are indicated by different lowercase letters (one-way ANOVA, P < 0.05).

It has been shown that PAP1 is degraded by the COP1 ubiquitin E3 ligase (Maier et al. 2013). Thus, we analyzed gene expression and anthocyanin contents in the cop1 knockout mutant with a high accumulation of PAP1 protein. The expression patterns of anthocyanin biosynthetic genes in cop1 were remarkably similar to that in pap1-D or BL-treated Col-0 (Supplementary Fig. S7, A and B; Fig. 3, A and B; Fig. 2, C and D). Furthermore, cop1 showed increased anthocyanin accumulation in response to BL (Supplementary Fig. S7, C and D).

Additionally, we analyzed the tt8 knockout mutant lacking a TT8 bHLH transcription factor among the MBW components. Compared to wild-type Col-0, BR-induced anthocyanin accumulation reduced considerably in the tt8 mutant (Supplementary Fig. S8, A and B). In our RT-qPCR analysis of anthocyanin biosynthetic genes upon BL treatment in pap1 and tt8 mutants, it was revealed that BL-induced expression of anthocyanin biosynthetic genes is dependent on the MBW complex-mediated transcriptional regulation (Supplementary Fig. S8, C and D). Inversely, both pap1 and tt8 showed no significant difference compared to the wild type in BL-induced hypocotyl elongation and root growth inhibition (Supplementary Fig. S8, E and F), suggesting that PAP1 and TT8 are not involved in growth regulation by BR signaling.

BZR1 directly binds to the PAP1 and PAP2 promoters in vitro and in vivo

Based on the observation that the expression pattern of anthocyanin biosynthetic genes in BL-treated Col-0 or bzr1-1D was very similar to that in the PAP1-overexpressing line, pap1-D (Fig. 2, C and D; Fig. 3, A and B), we hypothesized that the BR-responsive transcription factor BZR1 might directly regulate the expression of PAP1. In promoter analysis of the PAP1 gene, we observed that the PAP1 promoter contains 3 E-boxes (CANNTG) and 1 G-box (CACGTG) (Fig. 4A). An electrophoretic mobility shift assay (EMSA) of MBP-BZR1 using 4 E/G-box probes (PAP1-1 to 4) revealed that BZR1 strongly binds to the PAP1-3 probe containing a G-box motif (Fig. 4B). Specific binding of BZR1 to the PAP1-3 probe was further confirmed by an EMSA using a mutated PAP1-3 probe (Fig. 4C).

Figure 4.

BZR1 directly regulates the expression of PAP1 and PAP2. A) Schematic diagram of the PAP1 promoter. The promoter regions used as probes for EMSA are numbered as 1 to 4. The DNA fragments amplified in ChIP-qPCR analysis are numbered I and II. B) EMSA for MBP-BZR1 binding to PAP1 probes. ³²P-labeled DNA probes (PAP1-1 to 4) were incubated with 0 ng (−), 50 ng (+), or 100 ng (++) of MBP-BZR1. C) EMSA using mutated DNA probe or competitor DNA. The mutated DNA probe was generated by adenine substitution of the corresponding cis-element. Mutated and wild-type probe are denoted as “MT” and “WT”, respectively. The unlabeled cold DNA probe (cold probe; X1, X10, and X100) was used as a competitor DNA. The increase of cold probe was described as a right-angled triangle. D) Transient gene expression assay driven by the PAP1 promoter (n = 3). Together with ProPAP1:LUC reporter, Pro35S:YFP (control), Pro35S:BZR1, or Pro35S:HY5 plasmid was co-transfected into Arabidopsis mesophyll protoplast. The relative luciferase activity was calculated by vector control which was set as 1. Error bars indicate standard errors. LUC, luciferase. E) ChIP-qPCR analysis of BZR1 binding to the PAP1 promoter. Chromatin from Col-0 and BZR1-myc-overexpressing plants were immunoprecipitated by anti-myc antibodies. rDNA was used as a negative control. Error bars indicate standard deviations. The data shown are representative of 3 independent experiments. F) Schematic diagram of the PAP2 promoter. The promoter regions used as probes for EMSA are numbered 1 to 5. The DNA fragments amplified in ChIP-qPCR analysis are numbered I to III. G) Transient gene expression assay driven by the PAP2 promoter (n = 3). Together with ProPAP2:LUC reporter, Pro35S:YFP (control), Pro35S:BZR1, or Pro35S:HY5 plasmid was co-transfected into Arabidopsis mesophyll protoplast. The relative luciferase activity was calculated by vector control which was set as 1. Error bars indicate standard errors. LUC, luciferase. H) ChIP-qPCR analysis of BZR1 binding to the PAP2 promoter. Chromatin from Col-0 and BZR1-myc overexpression plants were immunoprecipitated by anti-myc antibodies. rDNA was used as a negative control. Error bars indicate standard deviations. The data shown are representative of 3 independent experiments. All of the statistically significant differences in this figure are indicated by different lowercase letters (one-way ANOVA, P < 0.05).

To investigate whether BZR1 binds to the PAP1 promoter in vivo, we performed protoplast-based effector/reporter assays. Luciferase activity driven by the PAP1 promoter was increased significantly by the addition of BZR1, similar to HY5, known to promote the PAP1 gene expression (Fig. 4D). In vivo binding of BZR1 to the PAP1 promoter was further confirmed by ChIP-qPCR analysis. In Col-0 and BZR1-myc plants, the protein–DNA complex was immunoprecipitated using anti-myc antibodies, and the eluted DNA was validated using qPCR analysis of well-known BZR1 target genes (Supplementary Fig. S9A). In a ChIP-qPCR analysis of the PAP1 promoter, the PAP1-II region (−49 to −220 bp from ATG) including the PAP1-3 and PAP1-4 probes, but not the PAP1-I region including the PAP1-2 probe, was enriched significantly (Fig. 4E).

We also examined whether BZR1 binds to the PAP2 promoter in vitro and in vivo. The PAP2 promoter contains 8 E-boxes and 1 G-box (Fig. 4F). Similar to BZR1 binding to the PAP1 promoter region, BZR1 is firmly bound to the G-box-containing DNA region of the PAP2 promoter in vitro (Supplementary Fig. S9B). In addition, significant binding of BZR1 to the PAP2 promoter was confirmed by a protoplast transfection assay (Fig. 4G). Consistent with the result of EMSA, the PAP2-III region (−78 to −218 bp from ATG) including the PAP2-5 probe, but not the PAP2-I and PAP2-II regions, was significantly enriched in the ChIP-qPCR analysis (Fig. 4H), suggesting that BZR1 binds to the G-box-containing region of the PAP2 promoter in vivo.

BZR1 physically interacts with PAP1 and PAP2 protein in vitro and in vivo

Interestingly, we found that the anthocyanin accumulation caused by the pap1-D mutation was significantly reduced in bri1-301 and Sdet2 mutants, in which BR signaling and biosynthesis are disrupted, respectively (Fig. 5A). This strongly suggests that, in addition to BR-induced PAP1 gene expression, BR activity is required for anthocyanin production regulated by PAP1 proteins. Thus, we hypothesized that BR-regulated BZR1 physically interacts with the PAP1 protein and cooperatively regulates anthocyanin biosynthesis. GST-BZR1 successfully pulled down MBP-PAP1 but not MBP-YFP, MBP-TT8, and MBP-TTG1 in vitro (Fig. 5B). We also investigated whether other BZR1/BES1 family members interact with PAP1. An in vitro pull-down assay confirmed that all 6 BZR1/BES1 family members interacted with PAP1 (Supplementary Fig. S10A). Notably, we observed that deletion of the N-terminal 89 amino acids of BZR1, including its DNA binding domain, resulted in the inability of BZR1 to bind PAP1 (Supplementary Fig. S10B).

Figure 5.

BZR1 interacts with PAP1 in vitro and in vivo. A) Comparison of anthocyanin level in the shoots of 14-d-old Col-0, bri1-301, Sdet2, pap1-D, pap1-D bri1-301, and pap1-D Sdet2 (n = 15). Error bars indicate standard errors. The statistically significant differences are indicated by different lowercase letters (one-way ANOVA, P < 0.05). FW, fresh weight. B) In vitro interaction of BZR1 and PAP1. The indicated MBP-fused proteins were pulled down with GST-BZR1-bound agarose bead. Immunoblots were detected with anti-MBP and anti-GST antibodies. C) Semi-in vivo pull-down assay for BZR1 interaction of PAP1. MBP-PAP1, but not MBP-YFP and MBP-TT8, was pulled down with BZR1-myc immunoprecipitated from BZR1-myc-overexpressing plants. Immunoblots were detected by anti-MBP and anti-myc antibodies. D) BiFC assays of BZR1 and members of the MBW complex. The indicated constructs were co-transfected into Arabidopsis protoplast. The combination of TT8-cYFP and nYFP-PAP1 was used as a positive control. Scale bars indicate 10 μm. E) Co-IP assay between BZR1 and PAP1. Transgenic Arabidopsis plants expressing YFP-PAP1 or co-expressing BZR1-myc and YFP-PAP1 were immunoprecipitated by GFP-Trap. Immunoblots were detected with anti-myc and anti-GFP antibodies. pBZR1 and BZR1 indicate phosphorylated and dephosphorylated form of BZR1, respectively. F) Changes in the interaction of BZR1 and PAP1 upon BL treatment. Transgenic Arabidopsis plants expressing YFP-PAP1 or co-expressing BZR1-myc and YFP-PAP1 were treated with mock or 50 nm BL for 5 h. The protein extracts were immunoprecipitated by GFP-Trap, and immunoblots were detected with anti-myc and anti-GFP antibodies.

Consistent with in vitro pull-down assay results, MBP-PAP1 but not MBP-TT8 interacted with BZR1-myc immunoprecipitated using anti-myc antibodies from BZR1-myc plants (Fig. 5C). We further confirmed in vivo interaction of BZR1 with PAP1 in Arabidopsis protoplasts using bimolecular fluorescence complementation (BiFC) assays (Fig. 5D). Notably, fluorescence signals were also detected in the BiFC combination of BZR1 with TT8 or BZR1 with GL3 although they did not directly interact with each other in vitro (Fig. 5, B and D; Supplementary Fig. S10, C and D). These results might indicate an indirect interaction of BZR1 with TT8 or BZR1 with GL3 within the same protein complex. Interestingly, EGL3 did not exhibit interaction with BZR1 in either in vitro or in vivo (Supplementary Fig. S10, C and D).

In a co-immunoprecipitation (Co-IP) assay, dephosphorylated forms of BZR1-myc were co-immunoprecipitated with YFP-PAP1 by GFP-trap (Fig. 5E). When YFP-PAP1 BZR1-myc plants treated with mock or BL were compared in a Co-IP assay, it showed that PAP1 predominantly interacts with dephosphorylated BZR1 rather than phosphorylated forms (Fig. 5F). Similar to PAP1, PAP2 also interacted with BZR1 in vitro and in vivo (Supplementary Fig. S11, A and B). Moreover, it was revealed that PAP2 preferentially binds to dephosphorylated BZR1 in vivo (Supplementary Fig. S11C).

BZR1 binds to the promoters of anthocyanin biosynthetic genes

The interaction observed between PAP1 and BZR1 suggests the possibility that BZR1 could bind to the promoter region of the target genes for PAP1. Thus, we examined whether BZR1 binds to the promoters of well-known PAP1-regulated genes, TT8 and DFR. We found 1 G-box, 1 PCE core (CNCCAC) known as a PAP1-binding site (Dare et al. 2008), and 5 E-box motifs in the TT8 promoter. It was notable that the G-box and the PCE core are very close and partly overlapped on the TT8 promoter (Fig. 6A). The result obtained from an EMSA using the designed 5 probes (TT8-1 to 5) indicated that BZR1 prominently binds to the TT8-1 probe (−1,076 to −1,122 bp from ATG) containing an overlapping region of the G-box and the PCE core motif (Fig. 6B; Supplementary Fig. S12A). Moreover, the TT8-I promoter region, including the TT8-1 probe, was significantly enriched in the ChIP-qPCR analysis using BZR1-myc plants (Fig. 6, A and C). We also investigated whether BZR1 binds to the DFR promoter containing 2 E-boxes, 4 G-boxes, and 2 PCE core motifs. We designed the DNA probes (DFR-1 and DFR-2) for 2 regions where G-box and PCE core are dense (Supplementary Fig. S12B) and confirmed MBP-BZR1 binding in an EMSA (Supplementary Fig. S12, C and D).

Figure 6.

BZR1 directly binds to the TT8 and LDOX promoter. A) Schematic diagram of the TT8 promoter. The promoter regions used as probes for EMSA are numbered as 1 to 5. The DNA fragments amplified in ChIP-qPCR analysis are numbered I and II. B) EMSA for MBP-BZR1 binding to TT8 probes. ³²P-labeled DNA probes (TT8-1 to 5) were incubated with 0 ng (−), 50 ng (+), or 100 ng (++) of MBP-BZR1. C) ChIP-qPCR analysis of BZR1 binding to the TT8 promoter. Chromatin from Col-0 and BZR1-myc-overexpressing plants were immunoprecipitated by anti-myc antibodies. rDNA was used as a negative control. Error bars represent standard deviations. The data shown are representative of 3 independent experiments. The statistically significant differences are indicated by different lowercase letters (one-way ANOVA, P < 0.05). D) Schematic diagram of the LDOX promoter. The promoter regions used as probes for EMSA are numbered as 1 to 3. The DNA fragments amplified in ChIP-qPCR analysis are numbered as I and II. E) EMSA for MBP-BZR1 binding to LDOX probes. ³²P-labeled DNA probes (LDOX-1 to 3) were incubated with the 0 ng (−), 50 ng (+), or 100 ng (++) of MBP-BZR1. F) ChIP-qPCR analysis of BZR1 binding to the LDOX promoter. Chromatin from Col-0 and BZR1-myc-overexpressing plants were immunoprecipitated by anti-myc antibodies. rDNA was used as a negative control. Error bars represent standard deviations. The data shown are representative of 3 independent experiments. The statistically significant differences are indicated by different lowercase letters (one-way ANOVA, P < 0.05).

Notably, in both TT8 and DFR promoters, BZR1 bound to the overlapping region of the G-box and PCE core motif. This led us to investigate whether the functional interaction between BZR1 and PAP1 requires a PCE core as well as a G-box motif. Thus, we examined the DNA binding of BZR1 to the LDOX promoter, which contains a G-box but not PCE core (Fig. 6D). Among 2 E-boxes and 1 G-box, BZR1 predominantly bound to the DNA probe possessing the G-box motif in EMSA (Fig. 6E; Supplementary Fig. S12E). In addition, the in vivo binding of BZR1 to the LDOX promoter was confirmed by the ChIP-qPCR analysis (Fig. 6F). Our results suggested that a PCE core motif is not essential for BZR1-mediated transcriptional regulation of anthocyanin biosynthetic genes.

The functional interaction of BZR1 with PAP1/2 enhances the expression of anthocyanin biosynthetic genes

We also investigated whether the DNA binding activity of BZR1 is increased by the interaction with PAP1. As expected, MBP-BZR1-bound amylose beads more effectively pulled down the biotin-labeled TT8-1 probes in the presence of His-PAP1 (Fig. 7A). Correspondingly, in a protoplast transfection assay, BZR1 promoted the TT8 promoter-driven luciferase expression, further enhanced by the co-expression of PAP1 (Fig. 7B). Similar to the case of the TT8 promoter, BZR1 more effectively bound to the DFR promoter in the presence of PAP1 in vitro and in vivo (Supplementary Fig. S13). Similarly, BZR1 more strongly bound to the LDOX promoter in the presence of PAP1 both in vitro and in vivo (Fig. 7, C and D). Our results demonstrated that the functional interaction of BZR1 with PAP1 on a G-box motif enhances anthocyanin biosynthetic gene expression.

Figure 7.

The interaction of BZR1 and PAP1/2 enhances the expression of TT8 and LDOX.A) PAP1 promotes BZR1 binding to the TT8 promoter. The MBP-BZR1-bound amylose bead preincubated with mock or His-PAP1 was incubated with the biotin-labeled TT8-1 probe. The DNA binding affinity of BZR1 was determined by the measurement of light intensity derived from the chemiluminescence reaction with streptavidin-HRP antibodies (n = 3). The relative DNA binding was calculated by empty control (first panel) which was set as 1. B) Transient gene expression assay driven by the TT8 promoter (n = 3). Together with ProTT8:LUC reporter, Pro35S:YFP, Pro35S:BZR1, or Pro35S:PAP1 plasmid was co-transfected into Arabidopsis mesophyll protoplast. The relative luciferase activity was calculated by vector control which was set as 1. LUC, luciferase. C) PAP1 promotes BZR1 binding to the LDOX promoter. The MBP-BZR1-bound amylose bead preincubated with mock or His-PAP1 was incubated with the biotin-labeled LDOX-3. The DNA binding affinity of BZR1 was determined by the measurement of light intensity derived from the chemiluminescence reaction with streptavidin-HRP antibodies (n = 3). The relative DNA binding was calculated by empty control (first panel) which was set as 1. D) Transient gene expression assay driven by the LDOX promoter (n = 3). Together with ProLDOX:LUC reporter, Pro35S:YFP (control), Pro35S:BZR1, or Pro35S:PAP1 plasmid was co-transfected into Arabidopsis mesophyll protoplast. The relative luciferase activity was calculated by vector control which was set as 1. LUC, luciferase. E) A ChIP-reChIP analysis for target promoters co-occupied by PAP1 and BZR1. Chromatin from Col-0 and PAP1-YFP BZR1-myc were subsequently immunoprecipitated by GFP-Trap and anti-myc antibodies. The precipitated DNA was quantified using qPCR analysis. rDNA was used as a negative control. qPCR was performed to amplify the promoter regions for TT8-I and LDOX-II shown in Fig. 6, A and D. DNA enrichment was calculated as the ratio between PAP1-YFP BZR1-myc and Col-0. The data shown are representative of 3 independent experiments. F) and G) Transient gene expression assay driven by the TT8 promoter F), and the LDOX promoter G) (n = 3). Together with ProTT8:LUC or ProLDOX:LUC reporter, Pro35S:YFP (control), Pro35S:BZR1, or Pro35S:PAP2 plasmid was co-transfected into Arabidopsis mesophyll protoplast. The relative luciferase activity was calculated by vector control which was set as 1. LUC, luciferase. H) A ChIP-reChIP analysis for target promoters co-occupied by PAP2 and BZR1. Chromatin from Col-0 and PAP2-YFP BZR1-myc were subsequently immunoprecipitated by GFP-Trap and anti-myc antibodies. The precipitated DNA was quantified using qPCR analysis. rDNA was used as a negative control. qPCR was performed to amplify the promoter regions for TT8-I and LDOX-II shown in Fig. 6, A and D. DNA enrichment was calculated as the ratio between PAP2-YFP BZR1-myc and Col-0. The data shown are representative of 3 independent experiments. Error bars in ChIP-reChIP results represent standard deviations and error bars in other results indicate standard errors. All of the statistically significant differences in this figure are indicated by different lowercase letters (one-way ANOVA, P < 0.05).

We further investigated whether the functional interaction of BZR1 with PAP1 occurs at promoters of anthocyanin biosynthetic genes in plant cells. To test whether PAP1 and BZR1 co-occupy the same promoter regions in vivo, we performed a ChIP-reChIP assay. In transgenic plants co-expressing PAP1-YFP and BZR1-myc, PAP1-binding regions of the TT8 and LDOX promoters were successfully enriched by subsequent immunoprecipitation using GFP-trap and anti-myc antibodies (Fig. 7E). Our results indicate that BZR1 is present in the same functional complex to which PAP1 belongs.

In luciferase reporter assays aimed at exploring whether the interaction between PAP2 and BZR1 enhances DNA binding activity to the TT8 and LDOX promoters in vivo (Fig. 7, F and G), it was observed that the transcriptional activity of PAP2 was relatively modest when compared with that of PAP1. Nevertheless, luciferase activity driven by BZR1 was significantly increased by the addition of PAP2. This functional interaction of BZR1 binding to PAP2 was further validated through a ChIP-reChIP assay, utilizing transgenic plants co-expressing PAP2-YFP and BZR1-myc (Fig. 7H).

To elucidate the central role of BZR1 in directly regulating downstream gene expression and its interaction with PAP1, we conducted a comparison between wild-type BZR1 and a mutated variant harboring the E37D mutation. A previous study demonstrated that Glu 37 residue of BZR1 is essential for DNA binding of BZR1 (Nosaki et al. 2018). Thus, we cloned BZR1^E37D that lacks the ability to bind DNA. In vitro pull-down assay confirmed that BZR1^E37D had almost similar binding affinity to PAP1 compared with wild-type BZR1 (Fig. 8A). We also conducted a luciferase reporter assay to investigate how DNA binding activity of BZR1 is important to the functional interaction of BZR1 with PAP1 (Fig. 8B). The result indicated that BZR1^E37D without DNA binding activity failed to increase the LDOX promoter-driven luciferase expression even in the presence of PAP1. Our findings strongly suggest that direct binding of BZR1 to the promoters of anthocyanin biosynthetic genes plays a more prominent role than its interaction with PAP1 in BR-induced anthocyanin accumulation.

Figure 8.

The DNA binding affinity of BZR1 plays a critical role in enhancing the expression of anthocyanin biosynthesis genes. A) In vitro pull-down assay for PAP1 interaction with wild-type BZR1 or mutated BZR1. MBP-BZR1 and MBP-BZR1^E37D, but not MBP-YFP, were pulled down with GST-PAP1-bound agarose bead. Immunoblots were detected with anti-MBP and anti-GST antibodies. B) Transient gene expression assay driven by PAP1 in the presence of BZR1 or mutated BZR1 (n = 3). Together with ProLDOX:LUC reporter, Pro35S:YFP (control), Pro35S:BZR1, Pro35S:BZR1^E37D, or Pro35S:PAP1 plasmid was co-transfected into Arabidopsis mesophyll protoplast. The relative luciferase activity was calculated by vector control which was set as 1. Error bars indicate standard errors. The statistically significant differences are indicated by different lowercase letters (one-way ANOVA, P < 0.05). LUC, luciferase. C) Biotinylation of MBW complex by BZR1-mediated proximity labeling. The indicated TurboID (TbID)-fused proteins and myc-tagged MBW complex were co-expressed in N. benthamiana leaves. Biotinylated proteins were pulled down by streptavidin-conjugated agarose beads and then detected with anti-GFP and anti-myc antibodies. Asterisks indicate the full-length proteins.

To further investigate whether BZR1 functionally interacts with the MBW complex in vivo, we performed a TurboID-based proximity labeling assay (Kim et al. 2023). BZR1-YFP-TurboID or YFP-YFP-TurboID was co-expressed in Nicotiana benthamiana leaves along with members of the MBW complex (PAP1-myc, TT8-myc, and TTG1-myc). Following biotin treatment, biotinylated proteins were pulled down by streptavidin beads and then detected using anti-myc antibody. Immunoblot assays revealed that BZR1-YFP-TuboID, but not YFP-YFP-TurboID, biotinylated all 3 MBW complex proteins fused to the myc tag in vivo (Fig. 8C). Our results strongly suggest that BZR1 forms a protein complex with PAP1, TT8, and TTG1 in plant cells.

BR plays a critical role in promoting anthocyanin accumulation under low nitrogen conditions

Considering that anthocyanin accumulation in plants occurs in response to diverse abiotic stresses, we questioned whether BR is required to produce anthocyanin under certain stress conditions. We noted previous studies showing that BR signaling modulates nitrogen uptake and root foraging response under nitrogen-deficient conditions (Jia et al. 2019, 2020; Chai et al. 2022; Wang et al. 2023). Thus, we aimed to investigate whether BR signaling influences anthocyanin biosynthesis under low nitrogen conditions.

First, we examined anthocyanin accumulation induced by low nitrogen conditions in pap1, tt8, and ttg1 mutants. The anthocyanin content induced by low nitrogen conditions in all 3 mutants was lower than in the wild type, indicating that low nitrogen conditions promote anthocyanin production through MBW complex-mediated regulation (Fig. 9A). In contrast, when we measured the root growth of pap1, tt8, and ttg1 mutants under different nitrogen conditions, we observed no discernible difference or very subtle phenotypic variations in root growth among these mutants under nitrogen starvation (Supplementary Fig. S14A).

Figure 9.

BR signaling is required for anthocyanin accumulation under low nitrogen conditions. A) Comparison of anthocyanin levels in the shoots of 10-d-old wild types, pap1, tt8, and ttg1 under high nitrogen (HN, 10 mm KNO₃) or low nitrogen (LN, 1 mm KNO₃) conditions (n = 15). Significant differences were determined by Student's t-test (n.s. no significant difference, and *P < 0.05). FW, fresh weight. B) and C) Comparison of anthocyanin levels in the shoots of 10-d-old Col-0, bzr1-1D, and ProBZR1:bzr1-1D-CFPB), and Col-0, Sdet2, and bzr-qC) under high nitrogen or low nitrogen conditions (n = 15). FW, fresh weight. D) to F) The relative gene expression level of PAP1D), DFRE), and LDOXF) in the shoots of 10-d-old seedlings grown on HN or LN medium. The statistically significant differences are indicated by different lowercase letters (one-way ANOVA, P < 0.05). The data shown are representative of 2 independent experiments. G) The relative expression of BZR1 and its homolog genes in the shoots of 10-d-old Col-0 seedlings grown on HN or LN conditions. Significant differences from the wild type were determined by Student's t-test (n.s. no significant difference, and *P <0.05). The data shown are representative of 2 independent experiments. All RT-qPCR data were normalized to the expression of the PP2A gene. Error bars in RT-qPCR results represent standard deviations and error bars in other results indicate standard errors.

Next, we examined anthocyanin accumulation of BR mutants under different nitrogen conditions. Under high nitrogen conditions, both bzr1-1D and ProBZR1:bzr1-1D-CFP did not display significant differences in anthocyanin levels when compared to Col-0. However, they exhibited a remarkable increase in anthocyanin levels compared with Col-0 in low nitrogen conditions (Fig. 9B). Similarly, bes1-D was more sensitive to low nitrogen conditions in terms of anthocyanin accumulation (Supplementary Fig. S14B). In addition, the decline in anthocyanin levels for Sdet2 and bzr-q was considerably more pronounced under low nitrogen conditions than under high nitrogen conditions, whereas bzr1 and bes1 showed no phenotypic difference (Fig. 9C; Supplementary Fig. S14, B and C). Accordingly, the elevation in PAP1, DFR, and LDOX expression was more prominent under low nitrogen conditions in the bzr1-1D compared with Col-0, whereas it was diminished in Sdet2 and bzr-q (Fig. 9, D to F). Our results indicate that BR signaling is required for the expression of anthocyanin biosynthetic genes and anthocyanin accumulation under low nitrogen conditions.

Next, we examined whether low nitrogen conditions are able to modulate BR signaling. An RT-qPCR analysis indicated that the expression of BZR1 and its homologs (BEH2, BEH3, and BEH4) was significantly increased in response to low nitrogen conditions (Fig. 9G). In line with these data, overall protein levels and relative ratio of dephosphorylated forms of BZR1-CFP were significantly increased by the low nitrogen conditions (Supplementary Fig. S15). Our results suggest that nitrogen deficiency enhances the expression of BZR1 as well as PAP1, thereby intensifying the functional interaction between BZR1 and PAP1.

Discussion

This study demonstrates that BR promotes anthocyanin biosynthesis through BZR1-mediated dual regulation of PAP1/2 in Arabidopsis seedlings. BZR1 directly induces the gene expression of PAP1/2 and physically interacts with PAP1/2 to cooperatively regulate the expression of genes implicated in anthocyanin biosynthesis such as TT8 and LBGs (Fig. 10).

Figure 10.

Proposed model for BR-regulated anthocyanin accumulation through functional interaction of BZR1 and PAP1. BR activates BZR1, leading to the upregulation of PAP1 expression. Moreover, BZR1 directly interacts with PAP1 protein of the MBW complex to cooperatively regulate the expression of anthocyanin biosynthesis genes. TT8 and TTG1 may play a role in the functional interaction of BZR1 with PAP1. In particular, under low nitrogen (N) conditions, the expression of PAP1 and BZR1 is induced, facilitating their functional interaction to promote anthocyanin accumulation. Solid arrows represent regulations in normal conditions, whereas dashed arrows indicate regulations in low nitrogen conditions. Solid arrows and dashed arrows represent established and undefined processes, respectively.

BR promotes the expression of PAP1 and PAP2, key transcription factors for anthocyanin biosynthesis, in Arabidopsis seedlings

We demonstrated that BR positively regulates anthocyanin accumulation in Arabidopsis shoots. Anthocyanin levels were reduced in BR-deficient and -insensitive mutants, and PCZ-treated Col-0 plants, whereas exogenously applied BL and bikinin greatly increased the anthocyanin contents (Fig. 1; Fig. 2; Supplementary Fig. S2). In addition, BR promoted the expression of various anthocyanin biosynthetic genes by transcriptional regulation of BZR1/BES1 family members. In particular, the expression of anthocyanin biosynthetic genes induced by BR was very similar to the gene expression pattern in PAP1-overexpressing plants (Fig. 2; Fig. 3). We found that BZR1 directly binds to the PAP1 and PAP2 promoter and promotes the expression of PAP1 and PAP2 (Fig. 4). In the pap1 knockout mutant, the effect of BL on anthocyanin accumulation was abolished largely (Fig. 3E). Our results suggest that PAP1 is a key target gene for promoting BR-mediated anthocyanin biosynthesis.

BL treatment increased the expression of EBGs such as CHS, CHI, and F3H as well as LBGs. A similar trend was also observed in pap1-D. It seems that PAP1 can directly or indirectly regulate the expression of EBGs in addition to LBGs. In contrast, although BL treatment induces UF3GT gene expression both in Col-0 and bzr1-1D, the basal expression level of UF3GT in bzr1-1D was reduced, compared to that in Col-0 (Fig. 2C). In fact, anthocyanin levels in bzr1-1D were not increased in the mock condition (Fig. 2B). Unlike bzr1-1D, under mock conditions, the anthocyanin levels in the bes1-D mutant were elevated compared to those in Col-0 (Supplementary Fig. S3A). Interestingly, the expression of UF3GT was not reduced in bes1-D (Supplementary Fig. S3B). UF3GT transferring a glucosyl group to anthocyanidin catalyzes a rate-limiting final step of the anthocyanin biosynthetic pathway. Similar to the BZR1-mediated negative feedback regulation of BR signaling, the expression of UF3GT might be limited by a BZR1-mediated negative feedback loop, thereby modulating the homeostatic levels of anthocyanins in the mock conditions.

In our ChIP-qPCR analysis and protoplast luciferase assays, BZR1 exhibited a higher binding affinity to the promoters of TT8 and LDOX compared to the PAP1 promoter (Fig. 4, D and E; Fig. 6, C and F; Fig. 7, B and D), implying that BZR1 might play a pivotal role in directly regulating the expression of the TT8 and LDOX genes.

Cooperative regulation of anthocyanin production by functional interaction of BZR1 and PAP1

It should be noted that anthocyanin accumulation induced by the pap1-D gain-of-function mutation was reduced significantly in bri1-301 and Sdet2. This suggests that despite excessive PAP1 transcription, the PAP1 protein is not fully functional without BR activity (Fig. 5A). In other words, this supports the notion that BR signaling is required for the transcriptional regulation of anthocyanin biosynthetic genes by PAP1. We found that the functional interaction of BZR1 and PAP1/2 enhances anthocyanin biosynthesis. Both PAP1 and PAP2 directly interacted with BZR1 in vitro and in vivo, elevating its DNA binding activity (Fig. 5; Fig. 7).

Several lines of evidence indicate that BZR1 cooperates with PAP1 and is connected to the MBW protein complex to regulate the expression of anthocyanin biosynthetic genes. First, BZR1 binds to the promoter regions of PAP1-target genes such as TT8, DFR, and LDOX in vitro and in vivo. In addition, BZR1 binding to the promoters of PAP1-target genes was increased significantly in the presence of PAP1 (Fig. 7). Second, a G-box motif is required for the functional interaction of BZR1 with PAP1. In the cases of TT8 and DFR, BZR1 predominantly binds to the overlapping region of the G-box and PCE core on the promoter region. However, given that cooperative regulation of BZR1 and PAP1 was observed even in the promoter of LDOX with only a G-box motif, overlapping of G-box and PCE core does not seem to be essential. Indeed, it was known that PAP1-binding motifs are relatively variable in the target gene promoter (Dare et al. 2008). Third, our ChIP-reChIP analysis demonstrates that BZR1 functions with PAP1 within the same transcriptional regulatory complex (Fig. 7E). Fourth, BZR1 interacted with TT8 and GL3 in an in vivo BiFC assay but not in an in vitro pull-down assay (Fig. 5, B to D). In BiFC assays, the fluorescence signals observed between BZR1 and TT8, as well as between BZR1 and GL3, are attributed to the close physical proximity between BZR1 and the 2 bHLH proteins. This proximity is facilitated by BZR1 binding to endogenous PAP1/2, which in turn forms the MBW complex with TT8 or GL3 in Arabidopsis protoplasts. Finally, in a proximity labeling assay, BZR1-YFP-TurboID biotinylated all 3 MBW complex members (Fig. 8C). This implies that BZR1 not only directly binds to PAP1 but is also positioned in proximity to the MBW complex that includes PAP1. In addition, BR-induced anthocyanin accumulation was reduced significantly in the tt8 mutant (Supplementary Fig. S8, A and B). Taken together, our results indicate that BZR1-mediated anthocyanin production depends on the function of the intact MBW complex.

Considering the positive role of BZR1 in anthocyanin biosynthesis, BZR1 may also cooperate with other members of the MBW complex. We found that the DNA binding domain, including the bHLH motif, of BZR1 is essential for the interaction of BZR1 with PAP1 (Supplementary Fig. S10B). A previous study has shown that the R3 domain of PAP1 is required for its interaction with bHLH factors (TT8, GL3, and EGL3) (Zimmermann et al. 2004). In contrast, the N-terminal region of GL3, rather than its bHLH motif, is involved in the interaction with PAP1 (Zhang et al. 2003). This suggests that BZR1 and bHLH factors bind to different regions of PAP1, implying that BZR1 might not compete with other bHLH proteins in forming the MBW complex.

In a previous study, it has been shown that JA promotes anthocyanin biosynthesis through the suppression of the JAZ repressor inhibiting MBW complex formation (Qi et al. 2011). In addition, it appeared that BR enhances JA-mediated anthocyanin biosynthesis in Arabidopsis seedlings (Peng et al. 2011). JA-induced expression of anthocyanin biosynthetic genes is greatly reduced in BR mutants, dwarf 4 (dwf4) and bri1. Our results suggest that the MBW complexes containing PAP1/2 cannot be sufficiently activated by JA in dwf4 and bri1, where BZR1 exists present in an inactivated state.

Hormonal regulation of anthocyanin biosynthesis under stress conditions

Anthocyanins in vegetative tissues are accumulated in response to various environmental stress conditions, which are often mediated by phytohormones (Steyn et al. 2002; Landi et al. 2015). High light intensity leads to the generation of reactive oxygen stress, which, in turn, stimulates the biosynthesis of abscisic acid (ABA) in the tea plant (Camellia sinensis) (Gao et al. 2023). ABA promotes the biosynthesis of anthocyanins and their transport into vacuoles, resulting in the leaves turning purple. In addition, ABA contributes to anthocyanin accumulation in Arabidopsis under low phosphorus conditions (Lei et al. 2022).

Notably, nitrogen deficiency significantly enhances the accumulation of anthocyanins in various plant species. Ethylene-insensitive Arabidopsis mutants exhibit reduced anthocyanin contents when subjected to low nitrogen conditions (Ma et al. 2023). Further, DELLAs, known as negative regulators of GA signaling, play a positive role in anthocyanin accumulation under low nitrogen conditions by direct binding with PAP1 (Zhang et al. 2017).

We observed BR-induced anthocyanin accumulation under normal conditions. Even in the absence of extreme stress conditions, BR not only promotes plant growth and development but also appears to induce the anthocyanin accumulation, thereby reducing incidental oxidative stress within cells during these processes. In addition, our results indicate that BR-induced anthocyanin accumulation becomes more evident in late seedlings and adult plants compared to early seedlings. Thus, BR-mediated anthocyanin production seems to confer resistance to stress that naturally arises with age and developmental progression under normal conditions. Nevertheless, it's worth noting that BR mutants, including bzr1-1D, exhibit a greater difference in anthocyanin contents compared to the wild type under low nitrogen conditions than under normal conditions. In addition, we showed in vivo binding of BZR1 with TT8 and GL3, but not with EGL3, leading to the combinatorial interaction within the MBW complex. Previous studies suggested that TT8 and GL3, but not EGL3, play a role in the accumulation of anthocyanins under stress conditions such as nitrogen starvation and low temperature (Feyissa et al. 2009). Accordingly, the specific interaction between BZR1 and 2 bHLHs may contribute to strengthening the transcriptional activity of the MBW complex in low nitrogen conditions.

Our study demonstrates that BZR1/BES1 family members activated by BR are required for PAP1-induced anthocyanin biosynthesis in response to low nitrogen stress. We show that the expression of BZR1 and its homologs was increased by low nitrogen conditions, leading to an increase of functional interaction between BZR1/BES1 family members and PAP1 (Fig. 8; Fig. 9). It should be noted that in tomato (Solanum lycopersicum) leaves, dephosphorylated BZR1 accumulates during nitrogen starvation (Wang et al. 2019 ). This accumulation stimulates autophagosome formation, serving as a survival strategy against low nitrogen levels. Similarly, it has been shown that low nitrogen conditions increase BR biosynthesis, which in turn promotes root foraging and subsequent nitrogen accumulation (Jia et al. 2020). Consistently, a nitrogen deficiency leads to the upregulation of BES1 expression and promotes dephosphorylation of the BES1 protein in Arabidopsis roots (Chai et al. 2022). BES1 also inhibits the transcriptional repression of LBD37 on nitrate-responsive genes. Taken together, BZR1/BES1 members confer resistance to nitrogen starvation by promoting anthocyanin accumulation and enhanced root foraging.

Materials and methods

Plant materials and growth conditions

The cyp85a1a2 (cyp85a1 cyp85a2-1), Sdet2, bri1-116, bin2-1, bzr1-1D, and pap1-D (CS3884, Arabidopsis Biological Resource Center) mutants are in Arabidopsis (Arabidopsis thaliana) Col-0, and the bin2-3 bil1 bil2 mutant is in Wassilewskija (WS) Arabidopsis accession background. The pap1 mutant and its accession Nossen-0 (No-0) were a kind gift from Prof. Youn-Il Park. All transgenic lines were generated in the Col-0 background. The loss-of-function mutants for BZR1, BES1, and BEH2 were kindly provided by Prof. Ming-Yi Bai. The bzr1, bes1, and beh2 mutants contain 90 (46 to 135 bp), 144 (58 to 201 bp), and 192 bp (14 to 205 bp) of deletion within the gene, respectively (Li et al. 2020; Tian et al. 2022). The beh3 mutant was obtained from ABRC (SALK_017577C).

Arabidopsis seeds were sterilized with 70% (v/v) ethanol containing 0.025% (v/v) Triton X-100 and washed in 100% ethanol. Sterilized seeds were planted on ½ MS (Duchefa Biochemie, Haarlem, Netherlands) medium containing 1% (w/v) sucrose and 0.8% (w/v) phytoagar (Duchefa Biochemie). Depending on the purpose of the experiment, different concentrations of BL (Olchemim, Olomouc, Czech Republic), bikinin (Sigma-Aldrich, St. Louis, MO, USA), or PCZ were treated in ½ MS medium. Seeds planted on MS medium were cold stratified at 4 °C for 3 d and then grown in a growth chamber at 22 °C under long-day conditions (16 h light/8 h dark cycle). Plants were exposed to white light under high light conditions (160 μmol m⁻² s⁻¹).

To examine the anthocyanin accumulation induced by low nitrogen conditions, sterilized seeds were planted on the medium containing ½ MS basal salt mixture without nitrogen (M531; Phytotechnology Laboratories, Lenexa, KS, USA), 1% (w/v) sucrose, 0.8% (w/v) phytoagar (Duchefa Biochemie), and 10 mm KNO₃ (high nitrogen, HN) or 1 mm KNO₃ (low nitrogen, LN). Plates were cold stratified at 4 °C for 3 d and then grown in a growth chamber at 22 °C under long-day conditions (16 h light/8 h dark cycle) for 10 d. To analyze root growth under nitrogen starvation, 7-d seedlings grown on high nitrogen-containing MS medium [½ MS without nitrogen, 1% (w/v) sucrose, 0.8% (w/v) phytoagar, 9.4 mm KNO₃, and 10.3 mm NH₄NO₃] were transferred to HN or LN medium and grown for additional 3 d.

For proximity labeling and affinity purification assay, we used 5-wk-old N. benthamiana grown on the soil at 22 °C under long-day conditions.

Plasmid construction

Coding sequences or promoter regions (∼1.5 kb) were cloned into Gateway pENTR 3C Dual Selection Vector (Invitrogen, Carlsbad, CA, USA). Entry clones were subcloned into gateway-compatible binary vectors [pEarleygate104 (Earley et al. 2006), BiFC vectors (Gampala et al. 2007), pDEST15 (Invitrogen), pGreenII-0800-LUC (Hellens et al. 2005), pCAMBIA1390-4myc or 7myc (Wang et al. 2013), and 35S-YFP-TurboID vector (Kim et al. 2023)] by LR reaction kit (Invitrogen). For bacterial expression, coding sequence of PAP1 was subcloned into pMAL-c2x, gateway-compatible pMAL-c2x, or pET-28a vector. For BZR1/BES1 family members, their coding sequences of wild type or mutated form were subcloned into pMAL-c2x vector.

Measurement of anthocyanin contents

Anthocyanin contents were measured as previously described (Deikman and Hammer 1995; Li et al. 2016). Fourteen-d-old Arabidopsis shoots were weighed and boiled for 3 min in 1 mL of anthocyanin extraction buffer [18% (v/v) iso-propanol, 1% (v/v) HCl, and 81% (v/v) H₂O]. The samples were incubated overnight at room temperature in the dark. The extracts were centrifuged at 17,000 × g for 5 min, and the absorbance of the supernatants was measured at 535 and 650 nm. Relative anthocyanin content was calculated by A535 to A650 per gram of fresh weight.

RT-qPCR analysis

Total RNA was extracted from 50 mg of Arabidopsis shoots using a plant total RNA extraction kit (Macherey-Nagel, Dueren, Germany) and cDNA was synthesized from 3 μg of total RNA using ReverTra Ace qPCR RT Master Mix with gDNA remover (Toyobo, Osaka, Japan). Quantitative RT-PCR (RT-qPCR) analyses were performed with THUNDERBIRD Next SYBR qPCR Mix (Toyobo) on CFX96 Real-Time PCR Detection System (Bio-Rad Laboratories, Hercules, CA, USA). PP2A was used as internal control, and the sequences of primers used in RT-qPCR are listed in Supplementary Data Set 1.

Electrophoretic mobility shift assay

EMSA was performed as previously described (Song et al. 2022). DNA probes containing G-box or E-box from the promoter of each target gene were radiolabeled with α-³²P-TTP using a Klenow enzyme (New England BioLabs, Ipswich, MA, USA). The sequences of oligonucleotides used in probe synthesis are listed in Supplementary Data Set 1. The MBP-BZR1 proteins (0, 50, and 100 ng) were incubated with 2.3 pmole of labeled DNA probes in the binding buffer [50 mm Tris–HCl, pH 7.5, 250 mm NaCl, 2.5 mm DTT, 5 mm MgCl₂, and 10% (v/v) glycerol] at 30 °C for 1 h. The protein–DNA complexes were separated on 8% (w/v) acrylamide-TBE gel and visualized by autoradiography. For the competition assay, unlabeled competitor DNA (50- or 100-fold of labeled probes) was used in the reaction. All bases of the G-box motif were substituted with alanine to construct the mutated DNA probes.

Arabidopsis protoplast isolation and BiFC assay

Arabidopsis protoplasts isolation was performed as previously described (Yoo et al. 2007). In brief, the rosette leaves from 3- to 4-wk-old Col-0 plants were cut into 0.5 mm strips and soaked in an enzyme solution [20 mm MES, pH 5.7, 1.5% (w/v) cellulase R10, 0.4% (w/v) macerozyme R10, 0.4 m mannitol, 20 mm KCl, 10 mm CaCl₂, and 0.1% (w/v) BSA]. After 30 min of vacuum infiltration, the samples were incubated for an additional 3 h in the dark at room temperature. The undigested leaves were removed by filtration with Miracloth (Millipore, Burlington, MA, USA), and the protoplasts were harvested by centrifugation at 100 × g for 3 min. The protoplasts were resuspended in W5 solution (2 mm MES, pH 5.7, 154 mm NaCl, 125 mm CaCl₂, and 5 mm KCl) and rested on ice for 30 min. The W5 solution was carefully removed after the protoplasts settled down, and 5 mL of fresh MMG solution (4 mm MES, pH 5.7, 0.4 m mannitol, and 15 mm MgCl₂) was added and gently mixed. For plasmid transfection, 10 μg of each construct was added to 200 μL of protoplast. After the addition of an equal volume of PEG solution [40% (w/v) PEG4000, 0.2 m mannitol, and 100 mm CaCl₂], the protoplast/DNA mixtures were carefully inverted and incubated for 5 min. The transfection was stopped with the addition of 800 μL of WI solution (4 mm MES, pH 5.7, 0.5 m mannitol, and 20 mm KCl) and the transfected protoplasts were harvested by centrifugation at 100 × g for 2 min. The protoplast pellets were resuspended with W5 solution and incubated in a 6-well tissue culture plate.

For BiFC assays in Arabidopsis protoplast, each combination of BiFC constructs was transfected into the protoplast. After 16 h incubation in the dark, the protoplasts were exposed to white light for 1 h before confocal microscopy. The YFP fluorescence was detected using a confocal microscope (C2 Plus; Nikon, Tokyo, Japan) with excitation by a 488-nm argon laser and emission detection from 500 to 545 nm.

Transient luciferase reporter expression assay in Arabidopsis protoplasts

The BZR1, PAP1 and HY5 coding sequences were cloned in the pEarleygate104 vector and used as effector constructs. An empty pEarleygate104 vector was used as a negative control. Approximately 1.5 kb promoter sequences of each target gene were cloned into pGreenII 0800-LUC to generate the reporter constructs. The effector and reporter construct combinations were transfected into protoplasts, followed by re-suspension of the transfected protoplasts in W5 solution, which were then incubated under darkness for 16 h and exposed to white light for 1 h before harvest. Harvested protoplasts were lysed with protoplast lysis buffer [25 mm Tris-HCl, pH 7.8, 1 mm DTT, 1 mm EDTA, 10% (v/v) glycerol, and 1% (v/v) Triton X-100] and centrifuged at 1,000 × g for 2 min. The resulting supernatants were used in the following luciferase assay. The firefly and renilla luciferase activities were measured using the Dual-Luciferase Reporter Assay System (Promega, Madison, WA, USA) according to the manufacturer's instructions.

Chromatin immunoprecipitation assay

ChIP assays were performed as previously described (Gendrel et al. 2005; Ko et al. 2022) with some modifications. In brief, 2 g of wild-type Col-0 and Pro35S:BZR1-myc (Kim et al. 2019) seedlings grown for 14 d were treated with 50 nm BL for 3 h and cross-linked for 15 min in 1% (v/v) formaldehyde solution by vacuum infiltration. Cross-linking was stopped by adding glycine to a final concentration of 0.125 m and additional vacuum infiltration for 5 min.

The samples were ground in liquid nitrogen and extracted using 30 mL of extraction buffer 1 (0.4 m sucrose, 10 mm Tris–HCl, pH 8.0, 10 mm MgCl₂, 5 mm β-mercaptoethanol, 0.2 mm PMSF, and protease inhibitor cocktail). The protein solution filtered with Miracloth was centrifuged for 20 min at 3,000 × g at 4 °C and the pellet was resuspended thoroughly in 1 mL of extraction buffer 2 [0.25 m sucrose, 10 mm Tris–HCl, pH 8.0, 10 mm MgCl₂, 1% (v/v) Triton X-100, 5 mm β-mercaptoethanol, 0.2 mm PMSF, and protease inhibitor cocktail]. The resuspended pellet was centrifuged for 10 min at 12,000 × g at 4 °C, and the pellet was resuspended in 300 μL of extraction buffer 3 [1.7 m sucrose, 10 mm Tris–HCl, pH 8.0, 2 mm MgCl₂, 0.15% (v/v) Triton X-100, 5 mm β-mercaptoethanol, 0.2 mm PMSF, and protease inhibitor cocktail]. The suspended pellet was carefully layered over 300 μL of extraction buffer 3 and centrifuged for 1 h at 16,000 × g at 4 °C. The resulting chromatin pellet was resuspended in 300 μL of nuclei lysis buffer [50 mm Tris–HCl, pH 8.0, 10 mm EDTA, 1% (w/v) SDS, and protease inhibitor cocktail] and sheared by sonication to reduce the average DNA fragment size to around 500 bp, resulting in a smearing range of approximately 200 to 700 bp. The sonicated chromatin solution was centrifuged for 5 min at 12,000 × g at 4 °C, and the supernatant was diluted with ChIP dilution buffer [1.1% (v/v) Triton X-100, 1.2 mm EDTA, 16.7 mm Tris–HCl, pH 8.0, and 167 mm NaCl].

The resulting chromatin samples were immunoprecipitated by anti-myc antibodies (Cell Signaling Technology, Danvers, MA, USA, Cat. #9B11)-bound protein A agarose beads (Thermo Fisher Scientific, Waltham, MA, USA) or protein A agarose beads alone as a negative control. The chromatin-bound beads were washed sequentially with low-salt buffer [20 mm Tris–HCl, pH 8.0, 2 mm EDTA, 150 mm NaCl, 1% (v/v) Triton X-100, and 0.1% (w/v) SDS], high-salt buffer [20 mm Tris–HCl, pH 8.0, 2 mm EDTA, 500 mm NaCl, 1% (v/v) Triton X-100, and 0.1% (w/v) SDS], LiCl buffer [10 mm Tris–HCl, pH 8.0, 1 mm EDTA, 0.25 m LiCl, 1% (v/v) NP-40, and 1% (w/v) sodium deoxycholate], and TE buffer (10 mm Tris–HCl, pH 8.0, and 1 mm EDTA).

The immunoprecipitated chromatin-protein complexes were eluted with elution buffer [1% (w/v) SDS and 0.1 m NaHCO₃]. After reverse cross-linking, the proteins were degraded by proteinase K treatment at 45 °C for 1 h, and DNA was purified with a PCR purification kit (Labopass, Seoul, Korea) following the manufacturer's instructions. Purified DNA was analyzed by ChIP-qPCR. Data were normalized with the rDNA coding region and then normalized with no antibody control. The enrichment was calculated as the ratio between Pro35S:BZR1-myc and Col-0. The sequences of primers used in ChIP-qPCR are listed in Supplementary Data Set 1.

Purification of recombinant proteins

MBP, GST, or His-fused proteins were expressed in Escherichia colistrain BL21 (DE3). Cells were cultured at 37 °C to OD₆₀₀ 0.5. The cells were treated with 0.2 mm isopropyl-β-D-1-thiogalactopyranoside and incubated at 37 °C for 2 h (MBP-fusion protein) or 28 °C for 3 h (GST-fusion protein and His-fusion protein) to induce protein expression.

To purify proteins, harvested cells were resuspended in MBP binding buffer (20 mm Tris–HCl, pH 7.4, 200 mm NaCl, and 1 mm EDTA), GST binding buffer (1× PBS), or His binding buffer (50 mm sodium phosphate, pH 8.0, 300 mm NaCl, and 10 mm imidazole). Then, lysozyme was added to the resuspended cell at a final concentration of 0.2 mg/mL and sonicated with Vibra-Cell Ultrasonic Liquid Processor. Cell lysates were centrifuged at 12,000 × g for 20 min, and the supernatants were incubated with amylose resin (New England BioLabs), glutathione resin (GenScript), or HisPur Ni-NTA resin (Thermo Fisher Scientific) at 4 °C for 2 h. Protein-bound resins were loaded on poly-prep chromatography columns (Bio-Rad Laboratories) and washed 5 times with 5 mL of binding buffer. The MBP-, GST-, and His-fusion proteins were eluted with MBP (20 mm Tris–HCl, pH 7.4, 200 mm NaCl, 1 mm EDTA, and 10 mm maltose), GST (50 mm Tris–HCl, pH 8.0 and 5 mm glutathione), or His elution buffers (50 mm NaH₂PO₄, pH 8.0, 300 mm NaCl, and 250 mm imidazole), respectively. All recombinant proteins were concentrated with Amicon Ultra Centrifugal Filters (Millipore).

In vitro pull-down assay

For in vitro pull-down assay, GST-fusion protein-bound agarose beads were incubated with MBP-fusion proteins in the binding buffer [50 mm Tris–HCl, pH 7.5, 150 mm NaCl, 0.2% (v/v) NP-40, and 0.1 mg/mL BSA] for 2 h at 4 °C. The protein-bound beads were loaded onto Micro Bio-Spin columns (Bio-Rad Laboratories) and washed with wash buffer [50 mm Tris–HCl, pH 7.5, 100 mm NaCl, and 0.1% (v/v) NP-40] 3 times. The proteins were eluted with 2X SDS sample buffer [24 mm Tris–HCl, pH 6.8, 10% (v/v) glycerol, 0.8% (w/v) SDS, and 2% (v/v) β-mercaptoethanol] and subjected to SDS-PAGE. The protein interaction was analyzed by immunoblotting using anti-MBP (1:20,000, New England BioLabs, Cat. #E8032) and anti-GST antibodies (1:2,000, Santa Cruz Biotechnology, Santa Cruz, CA, USA, Cat. #sc-138).

Co-immunoprecipitation

Fourteen-d-old seedlings were ground in liquid nitrogen and resuspended in IP extraction buffer [50 mm Tris–HCl, pH 7.4, 150 mm NaCl, 10% (v/v) glycerol, 0.1% (v/v) NP-40, and protease inhibitor cocktail]. The protein extracts were centrifuged at 13,800 × g for 10 min and the supernatants were filtered with Miracloth. The resulting protein extracts were immunoprecipitated with GFP-Trap Agarose (Chromotek, Planegg-Martinsried, Germany) for 2 h. After washing 3 times with wash buffer [50 mm Tris–HCl, pH 7.4, 150 mm NaCl, 10% (v/v) glycerol, 0.02% (v/v) NP-40, and PIC], the bead-bound proteins were eluted with 2× SDS sample buffer. The proteins were subjected to SDS-PAGE and analyzed by immunoblotting using anti-GFP (1:5,000, TransGen Biotech, Beijing, China, Cat. #HT-801) and anti-myc antibodies (1:20,000, Cell Signaling Technology).

DNA binding assay

MBP-YFP or MBP-BZR1 protein was bound to the amylose resin in MBP binding buffer for 2 h at 4 °C. The protein-bound amylose resin or amylose resin alone was incubated without or with His-PAP1 in IP-100 buffer [50 mm Tris–HCl, pH 7.6, 100 mm KCl, 2 mm MgCl₂, 0.05% (v/v) NP-40, 0.1 mg/mL BSA, and protease inhibitor cocktail] for 30 min at room temperature. The protein-bound beads were washed with IP-100 buffer 2 times and incubated with 5′-biotinylated TT8-1 or DFR-1 probes in IP-100 buffer for 1 h at room temperature. The beads were washed with IP-100 buffer 2 times and washed with PBS containing 0.05% (v/v) Tween-20 (PBST). After blocking with PBST containing 0.5% (w/v) BSA for 30 min, the beads were incubated with streptavidin-HRP (0.5 μg/mL) for 1 h and then washed 3 times with PBST buffer. Chemiluminescence was measured using a Varioskan Flash system (Thermo Fisher Scientific).

DPBA staining

To observe flavonol accumulation of plant tissues, Arabidopsis seedlings were stained with DPBA (Sigma-Aldrich). Fourteen-day seedlings grown on ½ MS medium without or with 50 nm BL were stained with ethanol containing 0.25% (w/v) DPBA and 0.01% (v/v) Triton X-100 for 45 min and then washed with 70% (v/v) ethanol 2 times. Leaves of stained seedlings were observed by confocal microscope (C2 Plus; Nikon) with excitation by a 488-nm argon laser and emission detection from 500 to 545 nm.

DMACA staining

To visualize proanthocyanidin in the seed coat, Arabidopsis seeds were stained with DMACA (Sigma-Aldrich). Approximately 50 seeds from each genotype were stained with DMACA reagent [0.5% (w/v) DMACA in 3 m HCl/50% (v/v) methanol] for overnight in the dark and then washed with 70% (v/v) ethanol 3 times. Stained seeds were observed using a dissection microscope (SMZ745T; Nikon).

Proximity labeling and affinity purification of biotinylated proteins in N. benthamiana

The coding sequences of YFP and BZR1 were cloned into Gateway-compatible Pro35S:YFP-TurboID vector whereas the coding sequences of PAP1, TT8, and TTG1 were cloned into Gateway-compatible Pro35S:4myc-6xHis or Pro35S:7myc-6xHis vectors. The plasmids were transiently co-expressed in N. benthamiana leaves and treated with 50 µM biotin for 3 h.

For nuclei isolation from N. benthamiana, 1 g of N. benthamiana leaves was ground in liquid nitrogen and resuspended in lysis buffer [20 mm Tris pH 7.4, 250 mm sucrose, 25% (v/v) glycerol, 20 mm KCl, 2 mm EDTA, 2.5 mm MgCl₂, and 1 mm PMSF]. Lysate was filtered with miracloth and centrifuged at 1,500 × g for 10 min, 4 °C. Nuclei pellet was resuspended in NRBT buffer [20 mm Tris pH 7.4, 25% (v/v) glycerol, 2.5 mm MgCl₂, and 0.2% (v/v) Triton X-100] and centrifuged 1,500 × g for 10 min at 4 °C. The pellet was washed with NRBT buffer for 2 more times. Nuclei pellet was resuspended in NE-2 buffer [20 mm HEPES–KOH, pH 7.4, 2.5 mm MgCl₂, 250 mm NaCl, 20% (v/v) glycerol, 0.2% (v/v) Triton X-100, 0.2 mm EDTA, 1 mm DTT, and protease inhibitor cocktail] and subjected to sonication for 5 times with 3 s on and 2 s off interval. The sonicated sample was centrifuged at 12,000 × g for 10 min at 4 °C, and the supernatant was transferred to PD-10 desalting columns (Cytiva, Little Chalfont, UK) to remove free biotin.

The nuclei extract was binding with 30 µL of Pierce Streptavidin Magnetic Beads (Thermo Fisher Scientific) at 4 °C for 3 h. The beads were washed 3 times with NE-3 buffer [20 mm HEPES–KOH, pH 7.4, 2.5 mm MgCl₂, 150 mm NaCl, 20% (v/v) glycerol, 0.2% (v/v) Triton X-100, 0.2 mm EDTA, 1 mm DTT, and protease inhibitor cocktail], and Streptavidin-bound proteins were eluted with 50 µL 2× SDS sample buffer containing 0.4 m urea. The proteins were subjected to SDS-PAGE and analyzed by immunoblotting using anti-GFP (1:5,000, TransGen Biotech) and anti-myc antibodies (1:20,000, Cell Signaling Technology).

Statistical analyses

Statistical significances of data were determined by Student's t-test or one-way ANOVA with posthoc test. All of the statistical analyses results and posthoc test of ANOVA are specified in Supplementary Data Set 2.

Accession numbers

Sequence data in this article can be found in the Arabidopsis Information Resource or GenBank/EMBL databases under the following accession numbers: PAP1 (AT1G56650), PAP2 (AT1G66390), GL3 (AT5G41315), EGL3 (AT1G63650), TT8 (AT4G09820), TTG1 (AT5G24520), CHS (AT5G13930), CHI (AT3G55120), F3H (AT3G51240), F3′H (AT5G07990), DFR (AT5G42800), LDOX (AT4G22880), UF3GT (AT5G54060), PP2A (AT1G13320), CPD (AT5G05690), DWF4 (AT3G50660), PRE5 (AT3G28857), and PRE6 (AT1G26945).

Author contributions

S.H.L., S.H.K., and T.W.K. conceived the project, performed most of the experiments, and wrote the manuscript. T.K.P. carried out cloning experiments and generated some transgenic plants. Y.P.K. and J.W.L. supervised the study and assisted with the protein purification.

Supplementary data

The following materials are available in the online version of this article.

Supplementary Figure S1. Anthocyanin accumulation in adult plants of BR mutants.

Supplementary Figure S2. Anthocyanin accumulation by bikinin and BL treatment.

Supplementary Figure S3. BES1-regulated anthocyanin biosynthesis.

Supplementary Figure S4. Flavonol and proanthocyanidin contents regulated by BR.

Supplementary Figure S5. Anthocyanin levels in the bzr1 and bes1 mutant and RT-PCR analysis of BZR1/BES1 family expression in the bzr-q mutant.

Supplementary Figure S6. Anthocyanin accumulation in seedlings and at the adult stage of BZR1/BES1 family-overexpressing plants.

Supplementary Figure S7. BR-induced anthocyanin accumulation in Col-0 and cop1-6.

Supplementary Figure S8. BR-induced anthocyanin accumulation in Col-0 and tt8-6.

Supplementary Figure S9. ChIP-qPCR analysis of BZR1 target genes and BZR1 binding to the PAP2 promoter.

Supplementary Figure S10. In vitro binding of BZR1 with MBW complex members.

Supplementary Figure S11. PAP2 interacts with BZR1 in vitro and in vivo.

Supplementary Figure S12. Specific binding of BZR1 to the promoter regions of TT8, DFR and LDOX in vitro.

Supplementary Figure S13. PAP1 enhances BZR1 binding to the DFR promoter.

Supplementary Figure S14. Root growth and anthocyanin content in anthocyanin mutants and BR mutants under low nitrogen conditions.

Supplementary Figure S15. Nitrogen deficiency induces BZR1 accumulation.

Supplementary Data Set 1. List of primers used in this study.

Supplementary Data Set 2. Results of statistical analysis.

Funding

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (Ministry of Science and ICT, or Ministry of Education) (2021R1A2C1006617 and 2020R1A6A1A06046728 to T.W.K.).

Data availability

The data underlying this article are available in the article and in its online supplementary material.

Dive Curated Terms

The following phenotypic, genotypic, and functional terms are of significance to the work described in this paper:

• BES1 Gramene: AT1G19350

• BES1 Araport: AT1G19350

• MYBL2 Gramene: AT1G71030

• MYBL2 Araport: AT1G71030

• LBD37 Gramene: AT5G67420

• LBD37 Araport: AT5G67420

• LBD38 Gramene: AT3G49940

• LBD38 Araport: AT3G49940

• LBD39 Gramene: AT4G37540

• LBD39 Araport: AT4G37540

• MIZ1 Gramene: AT2G41660

• MIZ1 Araport: AT2G41660

• MYB111 Gramene: AT3G46130

• MYB111 Araport: AT3G46130

• MYB113 Gramene: AT1G66370

• MYB113 Araport: AT1G66370

• MYB114 Gramene: AT1G66380

• MYB114 Araport: AT1G66380

• COP1 Gramene: AT2G32950

• COP1 Araport: AT2G32950

• MAX2 Gramene: At2g42620

• MAX2 Araport: At2g42620

• PAP1 Gramene: AT1G56650

• PAP1 Araport: AT1G56650

• UF3GT Gramene: AT5G54060

• UF3GT Araport: AT5G54060

• PP2A Gramene: AT1G13320

• PP2A Araport: AT1G13320

• HY5 Gramene: AT5G11260

• HY5 Araport: AT5G11260

• SMXL6 Gramene: At1g07200

• SMXL6 Araport: At1g07200

• BEH3 Gramene: AT4G18890

• BEH3 Araport: AT4G18890