CRY1 inhibits COP1-mediated degradation of BIT1, a MYB transcription factor, to activate blue light-dependent gene expression in Arabidopsis

Authors


*(fax +82 54 279 5972; e-mail nam@postech.ac.kr).

Summary

Cryptochromes (CRY) are one of the two major classes of photoreceptors that perceive light stimuli in the UV-A to blue light region and they are involved in multiple aspects of plant growth and development. However, knowledge regarding their signaling transduction components and mechanisms remains limited. Here, we report that a MYB transcription factor Blue Insensitive Trait 1 (BIT1), plays an important role in controlling blue light responses. Hypocotyl growth responses indicate that BIT1 functions as a positive element in blue light signaling, since BIT1 antisense and knock-out lines show a reduced light response in blue light. BIT1 controls blue light-dependent expression of various genes such as PsbS, a member of the light-harvesting complex gene family. A transactivation assay showed that BIT1 regulates promoter activity of PsbS in a blue light-dependent manner and that it requires CRY1 for activation of the PsbS promoter. BIT1 undergoes degradation in darkness and CRY1 functions to stabilize BIT1 in a blue light-dependent manner. In contrast, COP1 binds to BIT1 and mediates its degradation. We propose that the PsbS promoter is activated in blue light via the blue light-dependent stabilization of BIT1 by CRY1, while in darkness BIT1 is degraded by COP1-mediated proteolysis.

Introduction

Plant fitness depends upon a variety of environmental cues, among which light provides one of the most important signals for regulation of plant growth and development (Casal and Yanovsky, 2005; Franklin et al., 2005). In higher plants such as Arabidopsis, the light information ranging from UV-A to far-red is perceived primarily by three classes of photoreceptors: the phytochromes, cryptochromes and phototropins (Chen et al., 2004). Phytochromes absorb red/far-red light and regulate various physiological processes, including seed germination, the shade-avoidance response, seedling development and flowering (Franklin et al., 2005). The Arabidopsis genome contains five phytochromes (PHYA to PHYE) and extensive studies have shown that plants utilize a complex network of cascades in phytochrome-mediated red/far-red light signaling (Casal and Yanovsky, 2005; Quail, 2002; Wang and Deng, 2003). Two classes of photoreceptors sense light in the UV-A/blue range, the phototropins and cryptochromes (Lin, 2000). Phototropins are blue light-activated receptor kinases that mediate phototropism, chloroplast movement and stomatal opening, in response to changes in the direction and/or intensity of light (Briggs and Christie, 2002). Cryptochromes (CRY) are photolyase-like blue light receptors that are present in bacteria, plants and animals (Cashmore, 2003). In plants, cryptochromes mediate various UV-A/blue light responses, including inhibition of hypocotyl elongation, enhancement of cotyledon expansion, anthocyanin accumulation, promotion of floral initiation and resetting of the circadian clock (Ahmad and Cashmore, 1993; Bagnall et al., 1996; Devlin and Kay, 2000; Guo et al., 1998; Mao et al., 2005; Mockler et al., 1999; Whippo and Hangarter, 2003). In this study, we examined a signaling cascade for the cryptochrome-mediated blue light response.

In Arabidopsis, the cryptochromes CRY1 and CRY2 regulate blue light-dependent growth, development and gene expression. CRY1 plays a more prominent role under high blue light irradiance, whereas CRY2 performs critical roles under low blue light irradiance (Lin et al., 1998). A third cryptochrome (CRY3) has been identified in Arabidopsis, and it functions as a single-stranded DNA photolyase (Kleine et al., 2003; Selby and Sancar, 2006). The blue light signal perceived by cryptochromes is then transmitted to downstream responses via various signaling components. In Arabidopsis, the blue light signaling components include SUB1, a Ca2+-binding protein (Guo et al., 2001), and AtPP7 (Moller et al., 2003), a Ser/Thr protein phosphatase. In blue light signaling, SUB1 and AtPP7 function as negative and positive regulators, respectively. SHB1 contains both SPX and EXS domains, and it functions in cryptochrome signaling (Kang and Ni, 2006). Transcription factors involved in blue light signaling include HYH, MYC2/ZBF1 and GBF1/ZBF2. HYH (Holm et al., 2002) is a homolog of HY5, and GBF1/ZBF2 (Mallappa et al., 2006) belongs to the basic leucine zipper transcription factors. MYC2/ZBF1 is a MYC transcription factor that binds to Z-boxes (Yadav et al., 2005). In addition to these components, blue light signaling is also controlled by general regulators of light signaling such as HY5 (Ang et al., 1998).

An intriguing mechanism has been suggested for cryptochrome signaling, based on the observation that cryptochromes bind COP1 (Li and Yang, 2007; Wang et al., 2001; Yang et al., 2001). COP1 is an E3 ubiquitin ligase that modulates photomorphogenesis (Boccalandro et al., 2004; Yi and Deng, 2005) by mediating light-dependent degradation of various transcription factors involved in light signaling. COP1-regulated transcription factors include HY5, HYH, LAF1 and HFR1 (Ang et al., 1998; Duek et al., 2004; Holm et al., 2002; Jang et al., 2005; Seo et al., 2003; Yang et al., 2005) and the stability of these transcription factors is regulated in a blue light-dependent manner via cryptochromes. It has been suggested that blue light-dependent conformational changes in cryptochromes are transmitted from the light-absorbing N-terminus to the COP1-binding C-terminus, suppressing COP1 activity in the CRY–COP1 complex and resulting in increased stability of the transcription factors due to inhibition of COP1-mediated proteolysis (Li and Yang, 2007).

In this study, we show that BIT1, a R2R3 MYB transcription factor, is a positive regulator of blue light-dependent development and gene expression. We further demonstrate that BIT1 is stabilized by CRY1 in blue light, but is degraded by COP1-mediated proteolysis in darkness, which leads to activation of a blue light-inducible promoter.

Results

BIT1 is a positive regulator of blue light signaling

The R2R3 MYB family is one of the largest families of transcription factors in the Arabidopsis genome and comprises more than 125 members (Stracke et al., 2001), many of which play crucial roles in plant development and growth. In an effort to identify the MYB transcription factors involved in photosignaling and leaf senescence, the two major interests in this laboratory, we systemically generated antisense transgenic lines in which DNA encoding the C-terminal portion of individual MYB transcription factor genes was expressed in the antisense direction using a vector previously developed for random antisense mutagenesis of Arabidopsis (Jun et al., 2002). The C-terminal portion of these proteins was chosen to avoid the highly conserved N-terminal MYB DNA-binding domain, which could non-specifically suppress the expression of related genes. In designing the antisense constructs, we chose sequences that did not show significant homology over a 20-nucleotide span in a BLAST search, using default parameters for most of the MYB genes. The antisense lines were examined for phenotypic variations in photosignaling and leaf senescence. In an early phase of screening, we identified an antisense line for MYB38 (At2g36890, Figure 1a) that exhibited reduced inhibition of hypocotyl growth, i.e. a long hypocotyl phenotype under white light. When we measured hypocotyl growth in the MYB38 antisense transgenic line in red, far-red and blue light spectra, to examine whether BIT1 functions under specific light conditions; we found the greatest change in inhibition of hypocotyl growth under blue light (Figure 1b). Since the antisense line conferred reduced sensitivity to blue light, MYB38 was renamed BIT1, for Blue Insensitive Trait 1. Accordingly, the antisense line was designated BIT1 AS-1. As the hypocotyl growth response depends upon light intensity, we examined whether the lesion in the antisense line could be observed over a range of light fluences. The data in Figure 1(c) show a distinguishable lesion in the blue light response in light fluences of 5 and 20 μmol m−2 sec−1, with hypocotyl lengths almost double those of the wild type under the latter condition. Under red light there was not such a clear response and the responses of the BIT1 AS-1 line did not differ significantly from those of the wild type under high-fluence light (10 μmol m−2 sec−1). At lower fluences (0.1 and 2 μmol m−2 sec−1) of red light, the antisense line showed a slight difference from the wild type, but this was not as pronounced as observed under blue light. The lesion in the BIT1 AS-1 line had a negligible effect on far-red light responses. These results suggest that BIT1 functions as a positive regulator in light signaling and that it exerts its effect primarily in blue light responses, with some negligible activity in red light and far-red light responses, respectively.

Figure 1.

 The BIT1 antisense line shows a lesion in blue light perception.
(a) The structure of the BIT1 gene and of its protein product. White boxes represent the 5′ and 3′ untranslated regions. Gray boxes indicate the region encoding the R2R3 type MYB domain. Thin lines represent introns. The region used for the generation of antisense lines is indicated by an arrow. The nucleotide number of the T-DNA insertion site is noted with respect to the start of the open reading frame.
(b) Hypocotyl growth responses of BIT1 AS-1 line seedlings under darkness, blue (10 μmol m−2 sec−1), red (10 μmol m−2 sec−1) and far-red (10 μmol m−2 sec−1) light. The growth responses of wild type (Col), phyA, phyB, cry1 and cry2 were included as controls. The cry1cry2 double mutant was included for comparison between the blue light responses of the cry1 and cry2 single mutants. The cry1cry2 double mutants were generated in a Ler background, and therefore it was used as a background control. Data are shown as means (n = 20) ± SD.
(c) Hypocotyl growth response of Col, BIT1 AS-1 and cry1 seedlings under various intensities of blue (B), red (R) and far-red (FR) light. This experiment was repeated three times. Data are shown as means (n = 20) ± SD. Relative length is hypocotyl length at the given light intensities relative to that in darkness.

The reduced blue light response phenotype was due to reduced expression of MYB38 in the antisense line (Figure 2a). As we designed the antisense construct to be gene specific, we detected downregulation of MYB38 expression in the dark-grown seedlings, but found no significant difference in the expression of closely (MYB87) or distantly related (LAF1) MYB transcription factors (Figure 2b). We also generated independent antisense transgenic lines using the same MYB38 antisense construct and were able to regenerate the BIT1 AS-1 phenotype (Figure 2c,d). We confirmed BIT1 as a positive regulator in blue light signaling by examining transgenic lines overexpressing a BIT1-GFP (green fluorescent protein) fusion transcript (Figure 2a) under the control of the cassava vein mosaic virus (CsVMV) promoter. These lines showed an enhanced blue response; the hypocotyl growth inhibition by blue light was more pronounced in the BIT1 GFP lines than in wild type (Figure 2c,d). Interestingly, BIT1 GFP lines also showed a hypersensitive response to red and far-red light (Figure S1).

Figure 2.

 BIT1 is a positive regulator in blue light signaling.
(a) The RT-PCR analysis of BIT1 expression in the wild type, two antisense lines (BIT1 AS-1, BIT1 AS-11) and two overexpression lines (BIT1 GFP-2, BIT1 GFP-16). In the overexpression lines, BIT1 was overexpressed as a BIT1-GFP fusion. One microgram of total cellular RNA was subjected to RT-PCR. ACT2 was used as a control.
(b) The RT-PCR analysis of expression of BIT1 and related MYB family genes (MYB87 and LAF1) in the wild type (Col) and BIT1 AS-1 lines. One microgram of total cellular RNA was subjected to RT-PCR. ACT2 was used as a control.
(c) Comparison of seedling phenotypes of wild type (Col), cry1, BIT1 AS-1, BIT1 AS-11, BIT1 GFP-2 and BIT1 GFP-16 lines, grown under blue light (10 μmol m−2 sec−1). The seedlings were grown for 5 days. Scale bar, 5 mm.
(d) Hypocotyl length comparisons of wild type (Col), cry1, BIT1 AS-1, BIT1 AS-11, BIT1 GFP-2 and BIT1 GFP-16 seedlings, grown under various intensities of blue light. The seedlings were grown for 5 days. Data are shown as means (n = 20) ± SD.
(e) Anthocyanin contents of wild type (Col), cry1, BIT1 AS-1, BIT1 AS-11, BIT1 GFP-2 and BIT1 GFP-16 seedlings grown for 5 days under continuous blue light (20 μmol m−2 sec−1). The anthocyanin content (A535–A650/g fresh weight) is shown as means (n = 3) ± SD.

In Arabidopsis seedlings, the accumulation of anthocyanin is induced by light as a protective measure against photo-oxidative damage. This response is exerted by blue light via cryptochromes (Ahmad et al., 1995) and by far-red light through phytochrome A (phyA; Hoecker et al., 1998). Here, we examined whether the anthocyanin accumulation response induced by blue light is mediated through BIT1. Compared with the wild type, seedlings grown under blue light conditions exhibited reduced and increased anthocyanin levels in the BIT1 antisense and BIT1-GFP overexpression lines, respectively (Figure 2e). This result shows that BIT1 functions as a positive regulator in anthocyanin accumulation, as well as in inhibition of hypocotyl growth.

It has been suggested that RAX2 functions in the formation of accessory side shoots during the development of inflorescences in Arabidopsis and it also encodes MYB38 (Müller et al., 2006). The mutation rax2-1 (rax2-1/bit1-1 in this paper) has a T-DNA insertion in the MYB38 gene (WS background, Figure 1a), and the rax2-1/bit1-1 line was reported to have no noticeable phenotype as a single mutation (Müller et al., 2006). We examined the light response of rax2-1/bit1-1 by measuring inhibition of hypocotyl growth under various fluences of blue light. As shown in Figure S2(a), rax2-1/bit1-1 exhibited noticeably reduced inhibition of hypocotyl growth under 10 μmol m−2 sec−1 of fluence and accumulated less anthocyanin under blue light than wild type (WS) seedlings (Figure S2b). These data support the proposition that BIT1 functions as a positive regulator in blue light signaling.

BIT1 activates the PsbS promoter in a blue light-dependent manner

The various physiological responses mediated by blue light are accompanied by changes in gene expression (Folta et al., 2003; Ma et al., 2001; Ohgishi et al., 2004). Observing that BIT1 mediates blue light signaling, we then attempted to reveal its underlying molecular mechanisms using an assay system for blue light-mediated gene regulation by BIT1. We first examined whether BIT1 regulates blue light-mediated gene expression by examining the expression of blue light-inducible genes (PsbS and CHS) using quantitative RT-PCR. As shown in Figure 3(a), induction of the PsbS and CHS genes was attenuated in the BIT1 antisense and knock-out lines. This result shows that BIT1 functions as a positive regulator in blue light-mediated expression of these genes.

Figure 3.

 BIT1 activates the PsbS promoter under blue light via CRY1.
(a) Quantitative RT-PCR analysis of the expression of blue light-induced genes (PsbS, and CHS) in the wild type and BIT1 AS-1 lines, as well as the WS (genetic background of the T-DNA insertion mutant) and rax2-1/bit1-1 lines. Seedlings were grown for 5 days in the dark before transfer to blue light (10 μmol m−2 sec−1). Total cellular RNA was extracted at the times indicated after the transfer to blue light. Each reaction was performed with 1 μg of RNA. The transcript levels shown are means (n = 3) ± SD, relative to those of ACT2.
(b) Schematic of the effector and reporter used in the mesophyll protoplast assay for BIT1 transactivation of the PsbS promoter.
(c) Blue light-dependent transactivation of the PsbS promoter by BIT1. Mesophyll protoplast cells isolated from wild type, BIT1 AS-1 and BIT1 GFP-2 lines were transfected with the reporter plasmid (PsbS-Luc) alone or co-transfected with the effector plasmid (BIT1/PsbS-Luc), and then incubated in the dark (black bar) or in 10 μmol m−2 sec−1 of blue light (white bar). The means (n = 2) ± SD of luciferase activity relative to that of an internal control plasmid (UBQ10-GUS) are shown with normalization to the levels of the wild type in the dark without the effector plasmid. The numbers shown on the white bars indicate the fold change of luciferase expression in blue light relative to that in darkness.
(d) CRY1 is required for transactivation of the PsbS promoter by BIT1. Mesophyll protoplast cells isolated from wild type, cry1 and cry2 plants were transfected with the reporter plasmid (PsbS-Luc) alone or co-transfected with the effector plasmid (BIT1/PsbS-Luc) and then incubated in darkness (black bar) or in 10 μmol m−2 sec−1 of blue light (white bar). The cry1cry2 double mutant was included as a comparison to the blue light response of the cry1 and cry2 single mutants. Ler was included as a background control for the cry1cry2 double mutants, since the double mutant was generated in the Ler background. The means (n = 2) ± SD of luciferase activity relative to that of an internal control plasmid (UBQ10-GUS) are shown with normalization to the level of wild type cells in darkness without the effector plasmid. The numbers shown on the white bars indicate the fold change of luciferase expression in blue light relative to that in darkness.

Among the genes tested, PsbS showed the clearest difference in transcript levels between dark and blue light conditions. For further investigation of the role of BIT1 in blue light-mediated gene expression, we performed a transactivation assay in Arabidopsis protoplasts using a luciferase (LUC) reporter gene driven by the PsbS promoter (Figure 3b). Luciferase activity driven by the PsbS promoter was induced almost two-fold under the blue light assay conditions (Figure 3c). The introduction of full-length BIT1 into wild type protoplasts led to an enhancement of blue light-inducible expression of PsbS-LUC. The degree of blue light induction of PsbS-LUC was reduced in the antisense line (BIT1 AS), and this effect was negated by the introduction of full-length BIT1. In overexpression lines (BIT1 GFP), blue light expression of PsbS-LUC became greater than that in the wild type. These results show that BIT1 is involved positively in blue light-mediated transcriptional activation of the PsbS promoter.

BIT1 activation of the PsbS promoter requires CRY1

Since cryptochromes are involved in blue light-mediated gene expression in Arabidopsis, we examined whether blue light-mediated activation of the PsbS promoter by BIT1 occurs through cryptochromes. We used a luciferase transactivation assay to detect PsbS-LUC activation in protoplasts of CRY1 and CRY2 null mutants (Figure 3d). Blue light activation of PsbS-LUC was almost completely abolished in the cry1 mutant and could not be rescued by the introduction of further copies of BIT1. This finding indicates that CRY1 is critical for activation of the PsbS promoter, regardless of the BIT1 available. In contrast, activation of the PsbS promoter in the cry2 mutant resulted in only a slight, if any, difference from the wild type. These results suggest that the blue light-mediated activation of the PsbS promoter by BIT1 is achieved primarily via CRY1.

BIT1 is stabilized by blue light and regulated by CRY1

How does CRY1 enable BIT1 to activate the PsbS promoter in a blue light-dependent manner? In a few transcription factors (e.g. HY5 and HFR1) involved in blue light responses, stability is regulated through the activity of cryptochromes (Duek et al., 2004; Osterlund et al., 2000; Vandenbussche et al., 2007). This led us to examine whether the stability of BIT1 is also regulated by blue light. The amount of hemagglutinin (HA)-tagged BIT1 decreased gradually during incubation of wild type protoplasts in the dark (Figure 4a) and was undetectable at 2 h. Treatment with MG132, a 26S proteasome inhibitor, effectively blocked degradation of BIT1-HA in the dark (Figure 4a). Illumination of the protoplasts with blue light effectively increased the stability of BIT1-HA (Figure 4b). However, BIT1-HA was not stabilized by red light exposure (Figure S3). Together, these results show that, in darkness, BIT1 is degraded through a proteasome-dependent pathway and that it is stabilized by blue light. In the previous assay, CRY1 was required for blue light-dependent transactivation of the PsbS promoter by BIT1 (Figure 3). This led us to examine whether or not CRY1 mediates stabilization of BIT1 by blue light, and if this in turn leads to activation of the PsbS promoter by blue light. For this test, we utilized transgenic plants that expressed BIT1-GFP in the wild type and cry1 mutant backgrounds. As shown in Figure 4(c), the blue light-dependent stabilization of BIT1-GFP was observed in the wild type but not in the cry1 mutant. This result demonstrates that blue light stabilizes BIT1 via CRY1 in planta. In transgenic plants, overexpression of BIT1-GFP leads to an enhanced blue light response in hypocotyl growth inhibition (Figures 4d and 2d), as well as to increased activation of the PsbS promoter (Figure 3c). However, in cry1/BIT1 GFP transgenic line, the hypocotyl growth inhibition response mediated by blue light was reduced dramatically (Figure 4d), exhibiting a correlation between the blue light-dependent stabilization of BIT1 by CRY1 and the hypocotyl growth inhibition response.

Figure 4.

 BIT1 is degraded by proteasomes in darkness and stabilized by CRY1 in blue light.
(a) Degradation of BIT1-HA in the dark and its stabilization by the proteasome inhibitor MG132. Wild type mesophyll protoplasts were transfected with the BIT1-HA fusion construct that produces hemagglutinin (HA)-tagged BIT1 protein. Transfectants were then incubated overnight without (DMSO) or with 10 μm of MG132 in the dark. The cells were then treated with 100 μm of cycloheximide and harvested at the times indicated for extraction of total cellular protein. BIT1-HA proteins were detected by western blot analysis using anti-HA antibody. Coomassie Brilliant Blue-stained RbcS protein (RBC) served as a protein loading control.
(b) Stabilization of BIT1-HA protein by blue light. Mesophyll protoplast cells of wild type plants were transfected with the BIT1-HA fusion construct and incubated overnight with a low concentration (1 μm) of MG132 in the dark. The cells were then treated with 100 μm of cycloheximide and either kept in the dark or transferred to blue light (10 μmol m−2 sec−1), followed by harvest at the times indicated for extraction of total cellular protein. BIT1-HA proteins were detected by western blot analysis using anti-HA antibody. Coomassie Brilliant Blue-stained RbcS protein (RBC) served as a protein loading control.
(c) CRY1 is required for blue light stabilization of BIT1. Seedlings of transgenic plants expressing BIT1-GFP in wild type (BIT1 GFP) or cry1 mutant (cry1/BIT1 GFP) backgrounds were first grown in darkness for 5 days and then transferred to blue light (10 μmol m−2 sec−1). Total cellular proteins were extracted at the times indicated and BIT1-GFP was detected by western blot analysis using anti-GFP antibody.
(d) CRY1 is required for hypocotyl growth responses of BIT1 GFP transgenic line in blue light. Seedlings of the BIT1 GFP transgenic line in wild type (BIT1 GFP) or cry1 mutant (cry1/BIT1 GFP) backgrounds were grown for 5 days in darkness (Dark) or in 10 μmol m−2 sec−1 of blue light (Blue). The wild type and cry1 mutant without the BIT1-GFP construct were included as controls. The actual hypocotyl lengths are shown as means (n = 20) ± SD.

Degradation of BIT1 is mediated by COP1, and BIT1 interacts with COP1

How does CRY1 stabilize BIT1 in a blue light-dependent manner? It has been demonstrated that cryptochromes bind to the E3 ubiquitin ligase COP1, and that COP1-bound cryptochromes inactivate COP1 in a blue light-dependent manner, thereby leading to the accumulation of transcription factors that mediate downstream blue light responses (Ang et al., 1998; Duek et al., 2004; Wang et al., 2001; Yang et al., 2001). Since BIT1 is degraded by proteasomes in darkness, as well as being stabilized by blue light, we examined whether the stability of BIT1 is under the control of COP1-mediated proteolysis. Under dark conditions, we found that BIT1 exhibited greater stability in cop1-4 mutant protoplasts than in wild type cells (Figure 5a), which suggests that COP1 is involved in the degradation of BIT1. We then examined whether or not BIT1 binds to COP1, as some transcription factors involved in blue light signaling have been found to do so. In a yeast two-hybrid assay, BIT1 was found to interact with COP1 (Figure 5b). As a transcription factor, BIT1 exhibited some autoactivation activity in the yeast two-hybrid assay. However, the interaction between BIT1 and COP1 resulted in a 2.3-fold increase in β-galactosidase activity relative to the control. Binding between BIT1 and COP1 was confirmed using a bimolecular fluorescence complementation (BiFC) assay between N-terminal yellow fluorescent protein (YFP)-fused BIT1 (YFPN-BIT1) and C-terminal YFP-fused COP1 (YFPC-COP1) (Figure 5c and Figure S4). Other COP1-interacting proteins (HY5, HYH, LAF1, HFR1 and STH2) have been reported to co-localize with COP1 nuclear speckles (Ang et al., 1998; Datta et al., 2007; Holm et al., 2002; Jang et al., 2005; Seo et al., 2003), and we also observed nuclear speckles in protoplasts expressing YFPN-BIT1 and YFPC-COP1. BIT1-GFP protein localized to the nucleus in Arabidopsis protoplasts and in the BIT1 GFP transgenic line (Figures S5 and S6). Furthermore, co-expression of COP1-HA and BIT1-GFP resulted in the formation of nuclear speckles containing BIT1-GFP (Figure S5), suggesting that COP1 recruits BIT1 to the nuclear speckles. Furthermore, we confirmed interaction between BIT1 and COP1 using an in vivo co-immunoprecipitation assay, utilizing protoplasts in which BIT1-HA was co-expressed with COP1-GFP. When a whole cell lysate from the protoplasts was co-immunoprecipitated with an agarose-conjugated monoclonal anti-GFP antibody (αm GFP), BIT1-HA was pulled down along with COP1-GFP (Figure 5d), suggesting that BIT1 interacts with COP1 in plant cells.

Figure 5.

 BIT1 binds to COP1 and is more stable in the cop1-4 mutant.
(a) Stabilization of BIT1 in the cop1-4 mutant. Mesophyll protoplasts from the wild type and cop1-4 mutant were transfected with the BIT1-HA fusion construct and incubated overnight with a low concentration (1 μm) of the proteasome inhibitor MG132 in darkness. The cells were then treated with 100 μm cycloheximide, and kept in the dark before harvest at the time points indicated for extraction of total cellular protein. BIT1-HA proteins were detected by western blotting using anti-hemagglutinin (HA) antibody. Coomassie Brilliant Blue-stained RbcS protein (RBC) served as a protein-loading control.
(b) Interaction between BIT1 and COP1 in the yeast two-hybrid system. Prey vector with (COP1) or without (vector alone) the COP1 insertion was introduced into yeast with the bait vector with (BIT1) or without (vector alone) the BIT1 insertion. Interactions were monitored quantitatively using the liquid β-galactosidase activity assay. The β-galactosidase activity is shown as an arbitrary unit for mean (n = 3) ± SD.
(c) Interaction between BIT1 and COP1 using a bimolecular fluorescence complementation (BiFC) assay. The BIT1 (YFPN-BIT1) and COP1 (YFPC-COP1) coding sequences were fused to sequences encoding either the N- or C-terminal fragments of yellow fluorescent protein (YFP), respectively. These constructs were co-transfected into mesophyll protoplasts and BiFC signals deriving from interactions between YFPN-BIT1 and YFPC-COP1 were observed in nuclear speckles (green). The merged image of the BiFC signal and 4′,6-diamidino-2 phenylindone (DAPI) staining is shown at the bottom (Merged), along with the differential interference contrast (DIC) image. Scale bars, 10 μm.
(d) Interaction between BIT1 and COP1 in an in vivo co-immunoprecipitation assay. BIT1-HA, GFP and COP1-GFP fusion proteins were expressed in the combinations indicated. Co-immunoprecipitation was performed using agarose-conjugated anti-GFP monoclonal antibody (αmGFP). BIT1-HA, GFP and COP1-GFP in the whole-cell lysates (Input) and in the pellet fraction (Output) were detected by immunoblot analysis with horseradish peroxidase (HRP)-conjugated anti-GFP (αmGFP-HRP) monoclonal antibodies and HRP-conjugated anti-HA (αmHA-HRP) monoclonal antibodies. Under these conditions, anti-HA (αmHA-HRP) monoclonal antibodies detected BIT1-HA at the expected molecular weight (BIT1-HA), as well as in a slowly moving protein band (*).

Discussion

Here, we have presented genetic, molecular and biochemical evidence which suggests that BIT1 functions as a positive regulator of blue light signaling in Arabidopsis. The BIT1 antisense and knock-out lines showed reduced blue light responses in inhibition of hypocotyl growth, accumulation of anthocyanin and gene expression. However, it is noted that BIT1 may not be involved solely in blue light responses, since the antisense lines showed a slight lesion in the red light response. Furthermore, BIT1-GFP overexpression lines showed enhanced photomorphogenesis under red and far-red light, as well as under blue light. Although ectopic expression often results in pleiotropic phenotypes, our results suggest a possible involvement of BIT1 in all of these light signaling pathways. A similar observation has been made in HFR1-GFP overexpression lines (Yang et al., 2005). Interestingly, BIT1-GFP overexpression lines displayed a slightly shorter hypocotyl phenotype in the dark, compared with that of the wild type (Figures 2d and 4d). Overexpression of other AtMYB genes has also been shown to lead to photomorphogenic phenotypes in the dark (Newman et al., 2004; Shin et al., 2002).

From the evidence in this paper, we propose the following mechanism for transduction by BIT1 of the blue light signal to downstream responses such as gene expression and hypocotyl growth. BIT1 is a transcriptional activator of blue light-regulated genes, and in the dark COP1 interacts with BIT1 leading to proteasome-mediated degradation, which prevents BIT1-mediated photoresponses. Under our experimental conditions, the blue light signal was perceived by CRY1; the light used was in the high-fluence range where CRY1 functions predominantly and the role of CRY2 is minimal. Upon blue light illumination, blue light-activated CRY1 inhibits the activity of COP1 (Li and Yang, 2007; Yi and Deng, 2005), which results in the stabilization of BIT1. The resulting increase in levels of BIT1 leads to increased responses to blue light.

The stabilization mechanism we suggest for BIT1 is highly reminiscent of the stability regulation of other transcription factors in blue light signaling (e.g. HY5 and HFR1). Our results suggest that the MYB transcription factor BIT1 should be included as an additional member of the COP1-regulated transcriptional factors. Further genetic and biochemical studies will be required to determine whether or not BIT1 is a substrate for COP1. Our observations support the notion that CRY1 provides a blue light-dependent stabilization of transcription factors in the CRY–COP1 complex and that this represents an important regulatory mechanism in blue light signaling.

Although the MYB transcription factors are known for their diverse roles in plants, only a few are known to play roles in light responses (Ballesteros et al., 2001; Newman et al., 2004; Shin et al., 2002). Our results indicate that the MYB transcription factor MYB38 is a positive regulator of blue light signaling. It is highly intriguing that previous reports suggested a role for BIT1/MYB38 in shoot branch formation (Müller et al., 2006). However, any connection between this function and blue light signaling remains unclear at this point. One possibility is that since MYB transcription factors can function as part of the transcriptional complex in plants (Ramsay and Glover, 2005), MYB38 may perform different functions, which depend upon its association with other transcriptional factors.

Many of the MYB family members are known to act as transcription factors that activate specific target genes containing MYB-binding sequences in their promoter regions. BIT1 has a typical R2R3 DNA-binding motif and the BIT1-GFP protein localized to the nucleus of Arabidopsis transgenic and protoplast cells (Figures S5 and S6). In addition, BIT1 acts as a transcription activator for the PsbS promoter (Figure 3c,d). These results lead to the possibility that BIT1 may also act as a DNA-binding transcription factor. Using an in vitro electromobility shift assay, we observed that BIT1 was able to bind specifically to a MYB-binding consensus sequence in the promoter region of PsbS (data not shown). However, we failed to identify in vivo binding between BIT1 and the PsbS promoter sequence using a chromatin immunoprecipitation assay (data not shown). Thus, it remains to be seen whether or not BIT1 binds directly to a target sequence or indirectly as part of a transcriptional complex.

The two Arabidopsis cryptochromes perform differential roles in blue light signaling. In our transactivation assays, blue light-dependent activation of the PsbS promoter required CRY1, and the role of CRY2 was minimal. An explanation for this observation would be that CRY2 may not be active under these experimental conditions, which consisted of a relatively short exposure (<4 h) to a high blue light fluence (>10 μmol m−2 sec−1). The differential roles played by CRY1 and CRY2 in blue light responses have been described previously (Kleine et al., 2007; Wade et al., 2001). Further analyses will be required to determine the possible involvement of CRY2 in regulating the stability and/or activity of BIT1.

Experimental procedures

Plant material and growth conditions

All plants were grown in an environmentally controlled growth room at 22°C with a 16-h/8-h light/dark cycle. For monochromatic light assays, seeds were cold-treated at 4°C for 3 days, sown on 0.1x MS minus sucrose plates, and then exposed to continuous white light for 12 h to induce uniform germination. The plates were transferred to monochromatic light conditions and incubated at 22°C for 5 days. Red, far-red and blue light was generated by light-emitting diode (LED) light sources at 470, 670 and 735 nm, respectively (Good Feeling, http://www.goodfeeling.co.kr). Fluence rates were measured with an Optical Power Meter (model 840; Newport, http://www.newport.com/). White light was supplied by fluorescent lamps. BIT1 AS-1, BIT1 AS-11, BIT1 GFP-2 and BIT1 GFP-16 were generated in the Col-0 ecotype background, whereas rax2-1/bit1-1 seeds (kindly provided by K. Theres, Max Planck Institute for Plant Breeding Research, Cologne, Germany) were from a WS background.

Generation of transgenic plants

To generated antisense lines of MYB38/BIT1, a 546-bp BIT1 cDNA fragment (349–894 from the translation initiation site of the BIT1 cDNA clone) was amplified by PCR using the primers 5′-GAATCCCTCATAGCCACCATGGCTCCTC-3′ and 5′-CTCGAGAGTAGTACAACATGAACTTATCCTCC-3′. The amplified fragment was cloned into pNB96 (Jun et al., 2002) using the EcoRI and XhoI restriction sites, resulting in insertion of the BIT1 sequence in the antisense orientation under the control of the dual 35S promoter. For generation of the BIT1-GFP overexpression transgenic lines, a full-length BIT1 cDNA (897 bp) was amplified by PCR from cDNA using the primers 5′-GAATTCATGGGTAGGGCTCCATGTTGTGA-3′ and 5′-CTCGAGGTAGTACAACATGAACTTATCCTCC-3′. The amplified fragments were then cloned into the pCsVMV-GFP vector under the control of the CsVMV (Verdaguer et al., 1996). Transformation of Arabidopsis was performed by the floral dip method (Clough and Bent, 1998) using Agrobacteriumtumefaciens strain AGL1, harboring the binary vectors.

Anthocyanin measurements

For anthocyanin determinations, 5-day-old blue light-grown seedlings were harvested and their fresh weight measured. Total anthocyanin was extracted from 50 seedlings overnight at 4°C using 300 μl of 1% (v/v) hydrochloric acid in methanol, as described previously (Kim et al., 2003). Relative anthocyanin concentrations were calculated as absorbance at 530 nm minus absorbance at 657 nm, and then normalized to the total fresh weight of tissue used in each sample.

Quantitative RT-PCR analysis

Quantitative RT-PCR amplifications and measurements were performed with an AB7300 Real Time PCR System (Applied Biosystems, http://www.appliedbiosystems.com/), according to the manufacturer’s instructions. Seedlings were harvested, frozen in liquid nitrogen and then ground under RNase-free conditions. The RNA was extracted using the Qiagen RNeasy plant mini kit (Qiagen, http://www.qiagen.com/), and treated with DNase I at 37°C for 30 min. The RNA was then reverse transcribed using the ImProme-II reverse transcription system (Promega, http://www.promega.com/), following the manufacturer’s instructions. Ten microliters of cDNA was diluted to 110 μl with water and 3 μl of the diluted cDNA was used for PCR reactions. For quantification of cDNA, RT-PCR was performed using the SYBR Premix Extaq (Takara, http://www.takara-bio.com/). The following PCR conditions were used: 94°C for 3 min, followed by 40 cycles of 94°C for 15 sec and 60°C for 34 sec. The following primer sets were designed to amplify products of ∼100 bp in length for the genes under examination: PsbS-F, 5′-CTCTTCAAACCCAAAACCAAAGCT-3′; PsbS-R, 5′-GCCTTTGTGAAACCAATCCCA-3′; CHS-F, 5′-CATCTTGGCTATTGGCACTGCT-3′; and CHS-R, 5′-CGGTCATGTGTTCACTGTTGGT-3′. The quality of the amplification was determined using the dissociation curve analysis of the AB7300 Real Time PCR system. Fold changes in expression levels were calculated using the comparative CT method, which was normalized against ACT2 (Hall et al., 2003) expression from the same samples. Each experiment was repeated twice with independent samples, and RT-PCR reactions were performed in triplicate for each of the samples.

Transactivation assay in Arabidopsis protoplasts

The full-length BIT1 cDNA was inserted into a plant expression vector containing two copies of the HA tag and driven by 35SC4PPDK promoter (Sheen, 1996). The 2-kb Arabidopsis PsbS promoter was fused to a plant expression vector containing the firefly luciferase gene (Hwang and Sheen, 2001). Mesophyll protoplasts were isolated from Arabidopsis and transfected as described previously, with some modifications (Hwang and Sheen, 2001). The PsbS-LUC reporter was transfected alone or co-transfected with the BIT1 effector into protoplasts isolated from wild type, BIT1 AS-1, BIT1 GFP-2, cry1, cry2 or cry1cry2 plants. Transfected protoplasts were incubated in darkness or under 10 μmol m−2 sec−1 of blue light for 4 h. The UBQ10-GUS construct was used as an internal control.

Immunoblot analysis

For immunoblots, proteins were resolved on SDS-PAGE gels containing 10% acrylamide, transferred to PVDF membranes (Millipore, http://www.millipore.com/) and detected with horseradish peroxidase (HRP)-conjugated monoclonal anti-HA (Roche, http://www.roche.com/) or HRP-conjugated monoclonal anti-GFP (Santa Cruz Biotechnology, http://www.scbt.com/) antibodies. The same membranes were stained with Coomassie Brilliant Blue G-250 (Sigma, http://www.sigmaaldrich.com/).

Yeast two-hybrid assay

The DupLEX-ATM system (OriGene Technologies, http://www.origene.com/) was used for the yeast two-hybrid assay. Full-length BIT1 and COP1 cDNAs were cloned into the pGilda bait and pJG4-5 prey vectors, which produced in-frame fusions with the LexA DNA-binding and B42 activation domain, respectively. The yeast strain EGY48 (MATa, trp1, his3, ura3, leu2::6 LexAop-LEU2) contains the lacZ reporter plasmid pSH18-34. This strain was transformed with the appropriate ‘bait’ and ‘prey’ plasmids and interactions were detected on 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-gal) medium. A β-galactosidase activity assay was performed on transformants, as described previously (Ryu et al., 2005).

Bimolecular fluorescence complementation (BiFC) assay

The full-length BIT1 and COP1 cDNAs were fused into plant expression vectors containing N- or C-terminal fragments, respectively, of YFP (YFPN and YFPC). Transfected protoplasts were incubated at 22°C in the dark and examined with a confocal laser scanning microscope (CLSM; Olympus FluoView FV1000, http://www.olympus-global.com/). Nuclei of protoplasts were visualized with 4′,6-diamidino-2 phenylindone (DAPI) staining.

In vivo co-immunoprecipitation assay

The full-length COP1 cDNA was fused to GFP encoding sequences controlled by the CsVMV promoter (Verdaguer et al., 1996). Arabidopsis mesophyll protoplasts were isolated from mature leaves of the wild type plants and transfected with BIT1-HA/GFP or BIT1-HA/COP1-GFP. Protoplasts were then supplemented with the proteasome inhibitor MG132 (10 μm) and incubated overnight at 22°C in the dark. Cells were harvested and solubilized in immunoprecipitation (IP) buffer [50 mm 2-amino-2-(hydroxymethyl)-1,3-propanediol (TRIS)-HCl (pH 7.5), 150 mm NaCl, 1 mm EDTA, 0.1% NP-40, 0.1% SDS, 1 mm phenylmethylsulphonyl fluoride (PMSF), 20 μm MG132 and protease inhibitor cocktail (Roche)]. The extracts were centrifuged at 12 000 g for 15 min at 4°C and then the supernatant was incubated with 3 μl of agarose-conjugated anti-GFP monoclonal antibody (Santa Cruz Biotechnology) for 2 h at 4°C, followed by recentrifugation. The pellet fraction was washed four times with IP buffer and protein samples were separated on 10% SDS-PAGE gels, transferred to polyvinylidene fluoride (PVDF) membranes, and detected with HRP-conjugated monoclonal anti-HA (Roche) or HRP-conjugated monoclonal anti-GFP (Santa Cruz Biotechnology) antibodies.

Acknowledgements

We thank Klaus Theres for the rax2-1 seeds, Ildoo Hwang for the BiFC vector and the Arabidopsis Biological Resource Center (ABRC) for the Arabidopsis mutants. This work was supported by a Korea Science and Engineering Foundation (KOSEF) grant funded by the Korea government (MOST, no. R15-2004-033-05002-0) and by Crop Functional Genomics Frontier Research Program grants (CG3132).

Ancillary