The Regulation of DWARF4 Expression Is Likely a Critical Mechanism in Maintaining the Homeostasis of Bioactive Brassinosteroids in Arabidopsis

DWF4 Is Localized to the Endoplasmic Reticulum and Is Rarely Expressed in Specific Tissues

Previously, it has been shown that BR biosynthetic enzyme DWF1 is located in the endomembrane system (Klahre et al., 1998). In addition, eukaryotic cytochrome P450 enzymes are generally located in the endoplasmic reticulum (ER; Schuler, 1996). Thus it is possible that the rate-determining enzyme in BR biosynthesis is targeted to a similar location to perform steroid C-22α hydroxylation. To determine the subcellular localization of DWF4, a full-length cDNA of DWF4 was fused to green fluorescent protein (GFP) in frame and transfected into mesophyll protoplast cells of Arabidopsis and visualized using confocal laser scanning microscopy (Fig. 1A). To express the DWF4 gene in protoplasts, we first used the cauliflower mosaic virus 35S promoter. However, we could hardly see both the GFP image and proteins in western-blot analysis (data not shown). This suggests that the protein is either degraded rapidly or that the GFP-DWF4 transcript has a low translational efficiency. Thus, we switched to a known stronger promoter, Cassava vein mosaic virus (CvMV), resulting in the signal shown in Figure 1. ER-specific localization of DWF4 was shown by examining colocalization with the ER-targeted binding protein (BiP; Koizumi, 1996; Fig. 1, C and D). Subcellular parts that emit green and red fluorescences overlapped to display yellowish color and reported that GFP-DWF4 was primarily positioned in the organelle where BiP accumulates (Fig. 1D).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Localization of DWF4 proteins in protoplasts isolated from Arabidopsis leaf mesophyll cells. A, CvMV:GFP-DWF4 construct was cotransfected with 35S:BiP-RFP in protoplasts and visualized for GFP (A) and RFP (C). Protoplasts observed with light microscope (B) and the merge of the GFP-DWF4 and BiP:RFP is shown in D.

Next, to examine the spatial and temporal expression pattern of DWF4, RNA gel-blot analysis with tissue-specific RNA was performed. Figure 2A illustrates that DWF4 expression is relatively stronger in tissues such as shoot apex and flower (SAF), roots, and dark-grown seedlings. Axillary buds and undifferentiated calli also showed expression (Fig. 2A). These DWF4-expressing tissues represent plant parts that participate in active growth by cell division and/or expansion. However, when compared to the immediate next-step enzyme, CPD, overall expression level is very low even in actively growing tissues (Fig. 2B).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

RNA gel-blot analysis for tissue-specific expression of DWF4. The DWF4 expression is restricted to actively growing tissues. A, RNA gel-blot analysis using total RNA extracted from various tissues of adult plant, dark-grown (3-d-old) seedlings, and callus. Overall, DWF4 transcript level is rarely detectable, but highest in actively growing tissues such as SAF, roots, and axillary bud. Lanes are denoted with Axil for axillary bud; SAF for shoot apical meristem and flower; Siliq for silique; Pedic for pedicel; Roset for Rosette leaves; and Dark for dark-grown seedling. B, The expression level of CPD and DWF4 analyzed by a Gene Atlas tool using microarray data available at Genevestigator (http://www.genevestigator.ethz.ch/; Zimmermann et al., 2004).

Tissues Identified by DWF4:GUS Histochemistry Positively Correlate with RNA Gel-Blot Analysis Patterns

Tissue-specific expression of DWF4 in RNA gel-blot analysis was confirmed by the DWF4:GUS reporter system. Two different reporter constructs that include 1,132 bp or 180 bp upstream from the AUG translation start codon, respectively, were made. A select line number 6 stably and consistently expresses the UidA gene in confined tissues over several generations. Figure 3 displays the spatial and temporal expression pattern of GUS stain in DWF4:GUS seedlings (Fig. 3, A and D). The GUS stain was detected as early as the embryo stage (Fig. 3A). When a silique was fully elongated, a mature embryo was dissected out and stained. Cotyledon margins and radicles were clearly GUS positive (Fig. 3A). To examine developmental changes, the seeds were germinated and grown for different times before being stained. Staining around the cotyledon margins remained until 1 d after germination (DAG; Fig. 3B). In addition, elongating hypocotyls and rudimentary collet tissue (joint tissues of roots and hypocotyls) were also noticeably GUS positive (Fig. 3B). The GUS stain in 3-d-old seedlings was localized to root tips, collets, and emerging leaves of shoot apices, leaving the cotyledonal margins only weakly stained (Fig. 3C).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

GUS histochemical assay in transgenic lines harboring 1,132-bp DNA fragment of DWF4 promoter. The DWF4:GUS expression pattern is positively correlated with the RNA gel-blot analysis. A, Embryos opened up from fully elongated siliques. B, Light-grown seedling at 1 DAG. C, Light-grown seedling at 3 DAG. D, Dark-grown seedling at 3 DAG. E to G, Developing flower and pedicel. H, Shoot apical meristem. I, Silique. J, Lateral root primordia. K, Pedicel. L, Stem and developing flower. M, Collet (joint tissues of shoots and roots). N, Mature leaf of 4-week-old adult plant. Unit bars in A and B = 0.5 mm, C and D = 1 mm, E = 2 mm, and F = 5 mm. Unit bar in E is applicable to F through N, but 1 mm in H.

The GUS staining patterns of the dark-grown seedlings differed from those of the light-grown ones (Fig. 3D). The GUS stain was detectable throughout the cotyledons, suggesting that light participates in controlling the spatial expression of DWF4. In addition to the root tip and collet, the elongating zone of the hypocotyls was also GUS positive in the dark (Fig. 3D; additional figures shown later).

To further examine the developmental regulation of the DWF4 gene expression, plant parts of 4-week-old adult plants were subjected to GUS histochemical assay. In general, actively growing tissues including shoot tips, joints between primary and secondary inflorescences, axillary buds, collets, root tips, and lateral root primordia were clearly stained (Figs. 3 and 4). When a GUS-stained flower was more closely examined, proximal parts of both filaments and gynoecia as well as the junction between stigma and valve were stained (Fig. 3, E–G, arrowheads). Root tips and primordia of lateral roots were all strongly stained (Fig. 3K). In addition, proximal parts of the pedicels were also GUS positive (Fig. 3, E and J, arrows) as early as they are formed in the shoot apices (Fig. 3, E and L, arrows), and remain stained until the siliques mature (Fig. 3J, arrow). GUS stain was not detectable in mature leaves and internodes (Fig. 3, L and N). The collet tissues that possess many emerging primordia of inflorescences displayed the greatest GUS stains (Fig. 3M). Emerging leaf primordia at the shoot tips were first stained throughout the blade, then the stain localized to the expanding hydathodes (Fig. 3, H and L). The axillary buds, shoot apices, roots, and dark-grown tissues that were found to possess DWF4 transcripts in the RNA gel-blot analysis (Fig. 2) were all clearly stained in the GUS histochemical analysis. This suggests that the 1,132-bp DWF4 promoter that was used in this study represents DWF4 gene expression.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

DWF4:GUS histochemical analysis in the light. DWF4:GUS seedlings in the different BR mutant backgrounds were grown for 7 d in the light before GUS staining. Sections A to D are mock-treated control and E to H are BL-treated seedlings. The strength of GUS stain increased in cpd-388 (B) and bin2/dwf12-1D (C), but decreased in bzr1-D (D) compared to control (A). DWF4:GUS was down-regulated in response to BL in the control (E), cpd-388 (F), and bzr1-D (H), but not in the dwf12-1D (G). Unit bar = 0.2 mm for roots, 0.5 mm for shoots.

The Level of DWF4:GUS Expression Reflects Exogenous and Endogenous Fluctuation of BR Levels

Previously, we found that DWF4 transcripts accumulate in both BR biosynthetic and insensitive mutants (Choe et al., 2001). To test if this is reflected in the histochemical pattern of DWF4:GUS plants, we crossed a DWF4:GUS line to a biosynthetic mutant cpd-388 and signaling mutants bri1-5 and bin2/dwf12-1D. As compared to wild type (Fig. 4A), GUS staining of 7-d-old seedlings was noticeably intensified in cpd-388 (Fig. 4, A and B) and the bin2/dwf12-1D (Fig. 4C) mutant background.

Dark-grown seedlings displayed a more severely altered localization pattern. Strong GUS staining was detected in cotyledons, root tips, and hypocotyls immediately underneath the shoot apices (Figs. 5A and 6C). Furthermore, the GUS activity in the bin2/dwf12-1D was detected throughout the roots and hypocotyl (Fig. 5C) of the dark-grown seedlings. Therefore, expression of the DWF4 gene has been disrupted in the bin2/dwf12-1D mutants in the dark.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

DWF4:GUS histochemical analysis in darkness. Three-day-old DWF4:GUS seedlings in different BR mutant background were subjected to GUS staining. Shown are mock-treated control (A, B, C, and D) and BL-treated plants (E, F, G, and H). Intensity of GUS stain increased in cpd-388 (B) and bin2/dwf12-1D (C), and decreased in bzr1-D (D) as compared to a control (A). In the bin2/dwf12-1D background, additional tissues that are not GUS positive in the wild-type background were also stained. DWF4:GUS was down-regulated in response to BL in the control (E), cpd-388 (F), and bzr1-D (H), but not in the bin2/dwf12-1D (G). Unit bar = 0.2 mm.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

RT-PCR and histochemical analysis of DWF4:GUS in bri1-5 background. BL-mediated DWF4:GUS reduction is dependent on functional BR receptor kinase, BRI1. A, RT-PCR (left) and DNA gel-blot analysis with RT-PCR product (RT-Southern, right). Total RNAs isolated from wild-type Columbia seedlings grown in the presence of mock, BL, or Brz were used as templates. B, RNA gel-blot analysis using total RNA isolated from DWF4:GUS (left) or bri1-5/DWF4:GUS (right). C, GUS staining in DWF4:GUS (left column) and in DWF4:GUS/bri1-5 (middle) in the mock, and in DWF4:GUS/ bri1-5 in the presence of BL (right).

Recently, the plant-specific nuclear factor BRASSINAZOLE RESISTANT1 (BZR1) has been shown to control CPD and DWF4 expression. A gain-of-function mutation of this gene displays phenotypes of constitutive BR signaling, and lowers CPD expression levels (Wang et al., 2002). To test if the mutation also down-regulates DWF4, the DWF4:GUS construct was transferred to bzr1-D by genetic crossing. As shown in Figures 4D and 5D, the overall intensity of the GUS activity is weaker (Fig. 5D) compared to the wild-type control (Fig. 5A). Exogenous application of epi-BL does not noticeably change the GUS intensity in bzr1-D background (Fig. 5H).

Furthermore, to test if the increased GUS activity in the dwarf mutants was attributable to decreased endogenous BR levels, we supplemented the plants with epi-BL before examining their GUS activity. When BL was exogenously supplied, DWF4:GUS activity in the cpd-388 was diminished relative to plants without BL treatment (Figs. 4, B and F, and 5, B and F). This GUS activity in BL-treated bin2/dwf12-1D was not significantly decreased (Figs. 4, C and G, and 5, C and G), suggesting that the increased staining in these mutants is due not to shortage of endogenous BRs but to lack of feedback down-regulation of the DWF4 gene expression in this signaling mutant.

Modulation of the DWF4 expression levels by BL and Brz was confirmed by reverse transcription (RT)-PCR and RNA gel-blot analysis (Fig. 6, A and B). RT-PCR analysis of DWF4 transcripts indicates that BL reduces, whereas Brz increases, the steady-state level of the DWF4 mRNA (Fig. 6A). To test if the endogenous DWF4 promoter activity is consistent with the transgenic DWF4 promoter, we performed the same RT-PCR analysis for the GUS gene. As shown in Figure 6A, similar patterns of induction and suppression were observed. In addition, we examined gene expression of DWF4 by RNA gel-blot analysis. Brz efficiently increases DWF4 transcripts in the DWF4:GUS transgenic line (Fig. 6B, left). Furthermore, to test if the Brz-mediated induction of DWF4 requires functional BRI1, we did RNA gel-blot analysis with total RNA isolated from bri1-5/DWF4:GUS plants after treatment with BL or Brz (Fig. 6B, right). None of these treatments significantly changed DWF4 transcript levels in the bri1-5 mutant background (Fig. 6B). The increased level of DWF4 transcripts in the bri1-5 mutant background, which accumulates BRs compared to wild type, was visualized again at the tissue level. Unlike DWF4:GUS plants, DWF4:GUS staining in the bri1-5 mutant background was intensified at the margin of the cotyledons, junction between the shoot and root, and root tips in the light (Fig. 6C, top). In the dark, DWF4:GUS stains in the bri1-5 background were throughout the cotyledons, elongation zone of etiolated hypocotyls, and the root tips (Fig. 6C, bottom). Therefore, GUS staining was strongly intensified in the bri1-5 mutant background but the expression pattern was basically maintained at the tissue level (Fig. 6C). The stronger staining in bri1-5/DWF4:GUS may represent the sites of BL accumulation (Fig. 6C). As predicted based on RNA gel-blot analysis (Fig. 6B), BL-induced DWF4 feedback regulation disappeared in the bri1-5 mutant background (Fig. 6C), confirming the requirement of functional BRI1 in feedback regulation of DWF4 both in the light and dark. Intensified staining in bri1-5 background clarified where DWF4:GUS directs expression, which is low in wild-type background.

The Mechanism of DWF4 Transcriptional Regulation Is Distinct from That of CPD

Previously, Mathur et al. (1998) reported that transcription of CPD is repressed by BL and this process requires de novo synthesis of proteins, since BL-mediated CPD down-regulation was blocked by treatment with cyclohexamide (CHX). Interestingly, CHX treatment alone induced opposite reactions from CPD and DWF4 in that CHX treatment reduced CPD transcripts (Fig. 7, lane 1 and 2 on CPD), but increased DWF4 (Fig. 7, lane 1 and 2 on DWF4). For both CPD and DWF4, BL effectively down-regulated the transcripts, but BL + CHX attenuated the BL-mediated down-regulation. Similarly, Brz + CHX attenuated the Brz-mediated increase of CPD and DWF4 transcript levels. These data strongly suggest that CPD and DWF4 expression is differentially regulated: CPD expression is controlled primarily by transcriptional activation but DWF4 expression is controlled by strong inhibitory mechanism. However, involvement of specific positive/negative regulators of mRNA stability cannot be ruled out in either case.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7.

RT-PCR and GUS histochemical analysis of CHX effects on CPD and DWF4 transcription. After 2 h pretreatment with 100 μm of CHX, culture media were supplemented with 1 μm epi-BL or 1 μm Brz. After 6 h incubation, seedlings were harvested and used for RNA isolation. A, CHX treatment reduces the steady-state levels of the CPD transcripts (lanes 1 and 2 of CPD) and BL-mediated down-regulation of CPD is slightly attenuated by CHX (3 and 4 of CPD). Brz increases CPD level, but CHX inhibits this increase (5 and 6 of CPD). In contrast, CHX treatment increased DWF4 level (1 and 2 of DWF4), suggesting that repressors are de novo synthesized to maintain relatively low level of DWF4 compared to CPD (1 and 2 of DWF4), otherwise the CHX treatment results in a similar pattern as CPD. An rRNA row is shown as a loading control. B, The effect of CHX on DWF4 expression was confirmed using DWF4:GUS transgenic lines. GUS stains in the root tip (top sections) and lateral root (bottom sections) were significantly increased upon CHX treatment (2 and 4) compared to the controls (1 and 3). In contrast, CHX treatment combined with Brz (6) resulted in reduced level of staining relative to Brz treatment alone (5).

Endogenous BR Levels Are Increased in Actively Growing Tissues

DWF4 is an important flux-regulating enzyme in BR biosynthetic pathways, thus it is conceivable that its expression reflects the tissues of BR biosynthesis. To compare endogenous levels of BR biosynthetic intermediates in different tissues, rosettes, siliques, collets, stems, and shoot tips, were collected from 25-d-old Arabidopsis wild-type plants. Table I shows the endogenous levels of each of 13 biosynthetic intermediates in seven different tissues. In seedlings, roots show greater C28 sterol and BR contents than shoots. Of the five adult tissues, the shoot tip unequivocally has the greatest sterol level. However, differently from the sterol levels, collets have approximately 5-fold more BR than siliques. This enrichment of BR content is correlated with a greater level of DWF4:GUS expression in these tissues (Fig. 3M).

Successful detection of BL using gas chromatography/mass spectrometry generally requires more than 100 g of fresh tissue, which is labor intensive to collect from small tissues such as the Arabidopsis shoot tip. Thus we chose to analyze the tissues of the bri1-5 mutant to detect BL in different tissues, because it is known that bri1-5 accumulates BR biosynthetic intermediates and BL (Noguchi et al., 1999). Table II summarizes the endogenous levels of BRs in six different tissues of Arabidopsis bri1-5 mutants. The seeds and shoot tips possess significantly enriched amounts of BRs compared to other tissues.

DWF4 Is the Essential Enzyme That Determines the Tissue Specificity of BR Biosynthesis and Size of the BR Pool

Changes in sensitivity or concentration of phytohormones are two major mechanisms that trigger signal transduction pathways in specific tissues. Increased sensitivity can be achieved through production of receptor molecules or by elevating pools of bioactive signaling molecules. BRs are perceived by the plasma membrane-localized receptor BRI1 and trigger a specific signaling cascade. However, it has been shown that the expression of the BRI1 gene is not localized to specific tissues of Arabidopsis, but is ubiquitous (Friedrichsen et al., 2000). Furthermore, CYP85A2, the enzyme mediating the ultimate step in BL biosynthesis, is also broadly expressed in Arabidopsis tissues (Castle et al., 2005). In addition, CPD, an immediate next enzyme to DWF4, is also highly and ubiquitously expressed (Mathur et al., 1998; Fig. 2B). Thus, the enriched BR levels in the specific tissues may be due not to CY85A2 or CPD expression but to a localization of the rate-determining enzyme DWF4.

Previously, we found that only limited tissues in shoot tips of the dwf4-1 inflorescences respond to applied BRs by dramatic elongation of internodes (Choe et al., 1998). This suggests that the young competent cells clustered in the shoot tip respond to BR concentration and trigger BR-dependent elongation responses. To accomplish this localized response, a pool of bioactive BRs needs to be maintained in these specific tissues.

BR content is almost quadrupled in root tissues of seedlings (Table I), and the BL level in the shoot tip scores approximately 20-fold greater than in the stem of bri1-5. This clearly indicates that the pool of bioactive BRs accumulates in specific tissues, and localized response to BRs is induced by increased concentration of bioactive BRs in these tissues. In addition, the DWF4 gene expression pattern supports that the increased pool of BRs in these tissues may originate from de novo biosynthesis. The DWF4-expressing tissues revealed by the GUS histochemical analysis identified tissues with a greater amount of BR levels than those organs that have less DWF4 gene expression. Therefore, tissue specificity of BR biosynthesis, accumulation, and localized response of BRs is likely to be imparted by DWF4 expression.

BL-Induced Feedback Regulation of DWF4 Requires Functional BRI1

Several BR biosynthetic enzymes including DWF4, CPD, and CYP85 were shown to be feedback regulated upon BR treatment at the transcriptional level (Bancos et al., 2002; Shimada et al., 2003). Previously, feedback regulation of CYP90A, which is highly homologous to CYP85, was shown to be dependent on functional BRI1 since BL had no effect on the expression of CYP90A in the BR-insensitive cbb2 (allelic to bri1; Bancos et al., 2002). We found that DWF4:GUS was significantly up-regulated and was not down-regulated upon BL treatment in the bri1-5 and bin2/dwf12-1D mutant background (Figs. 4, 5, and 6). In addition, endogenous levels of BRs were significantly increased in bri1-5 tissues expressing high levels of DWF4 (Table II). Therefore, it is likely that BL-induced feedback regulation of DWF4 requires functional BRI1 and increased level of DWF4 expression is responsible for the increase of BR pools in this mutant.

BZR1 Represses DWF4 Expression

According to Wang et al. (2002), the brassinazole resistant mutation bzr1-D effectively down-regulates CPD transcript levels, suggesting that BZR1 plays a role as a repressor of BR biosynthetic gene CPD. Recently, BZR1 has been proposed as a transcriptional regulator that can bind directly to CPD and act as repressor (He et al., 2005). Interestingly, the expression pattern of BZR1 and DWF4 is mutually exclusive. The tissues just below the apical hook are noticeably stained for DWF4:GUS (Fig. 6C), whereas the corresponding regions in the bzr1-D mutant are not stained (Fig. 5D; Wang et al., 2002). Therefore, the DWF4:GUS staining pattern alternative to the BZR1 protein expression suggests that BZR1 plays an important role as a repressor of DWF4. Previously, Wang et al. (2002) showed that the bzr1-D protein is stable and present in these tissues where BZR1 is not. In our studies, DWF4:GUS activity was lower than corresponding control in the light- and dark-grown seedlings of bzr1-D. These results imply that BZR1 is a repressor of DWF4 expression.

BIN2/DWF12 Directs Localized Expression of DWF4

BIN2/DWF12 encodes a highly conserved Ser/Thr kinase that negatively regulates BR responses by phosphorylating two positive regulators, BZR1 and BZR2/BES1, to be targeted for degradation (Li and Nam, 2002; Zhao et al., 2002). In the dwf12-1D mutant background, DWF4:GUS was significantly up-regulated, especially in the dark. In addition, upon BL treatment, DWF4:GUS was not significantly down-regulated either in the light or dark. Since dwf12-1D has increased BR level (Choe et al., 2002), the up-regulation of DWF4:GUS in this mutant background is probably due to the lack of feedback regulation. Unlike CPD or bri1-5 mutant background, localization of DWF4:GUS was disrupted severely in the roots of dwf12-1D (Fig. 5, C and G). This is probably due to the enhanced degradation of repressor BZR1, whose phosphorylation and subsequent degradation is activated in dwf12-1D. The mislocalized pattern of the DWF4:GUS gene is obvious in the dwf12-1D background. For example, GUS stain seems to change the pattern from the vasculature (Fig. 5C, mock) to epidermal cell layer (Fig. 5G, BL treated) in the dark-grown dwf12-1D mutant seedlings. In addition, epidermal staining was also found in cotyledon margins of all four mutants tested (Figs. 5, F–H, and 6C). Taken together with more predominant ectopic expression of DWF4:GUS/dwf12-1D in the dark, light seems to be the key regulator involved in the regulation of DWF4 expression via regulation of BIN2/DWF12.

Regulation of DWF4 Transcription Is Distinct from That of CPD

GUS histochemical analysis patterns of DWF4 and CPD are biphasic; they have overlapping expression, but also distinct tissues. Like CPD, DWF4:GUS activity is seen in the emerging leaf primordia and margins of developed leaves (Mathur et al., 1998). However, differently from CPD, GUS activity driven by the DWF4 promoter is obvious in the dark-grown cotyledons, whereas light-grown cotyledons have only their margins and hydathodes stained (Fig. 3, A and H). In addition, DWF4:GUS expression was detected in root tips, whereas CPD:GUS was not (Mathur et al., 1998).

Regulation of CPD and DWF4 seems to be quite distinct since they have different tissue/organ specificity as revealed by this study and patterns identified by Mathur et al. (1998). In addition, the mechanisms of DWF4 and CPD gene expression are quite distinct at the transcriptional level, since CHX treatment increases DWF4 transcripts, but decreases CPD transcripts. This strongly suggests that CPD expression is regulated mainly by transcriptional activator but DWF4 expression is by strong inhibitory mechanism. Thus, inhibition of protein synthesis by CHX treatment resulted in opposite effects on the DWF4 and CPD gene transcription.

Involvement of BZR1 as a repressor and localized expression of the DWF4 gene was clearly demonstrated in this study. However, BZR1-mediated inhibition of DWF4 seems to be only one among various inhibition mechanisms since the transcript level of DWF4 is extremely low compared to that of CPD, although regulation of both genes are shown to be controlled by transcription repressor BZR1 (He et al., 2005). In addition, it has been shown that BZR1 binds more tightly to CPD than DWF4 for its inhibitory action. Therefore, involvement of other inhibitors of DWF4 cannot be excluded, as implicated by differential effects of CHX on DWF4 and CPD. Further studies should be focused on the identification of inhibitory or activating components that are involved in DWF4 transcriptional regulation.

In conclusion, unlike other genes involved in BR biosynthesis and signal transduction pathways, such as CPD, BL synthase (CYP85A2), and BRI1, DWF4 expression is tightly regulated in specific tissues of Arabidopsis. Its expression is feedback down-regulated by bioactive BRs. Reduction in endogenous BR levels in a cpd-388 mutant background or by treatment with Brz induces DWF4 expression, but exogenous application of BRs decreases DWF4 expression. Both transcriptional and posttranscriptional regulations of DWF4 are important means that determine the level of bioactive BR. In this study, we found that DWF4 protein is extremely labile; GFP fluorescence by transient expression in protoplasts through known strong promoter cauliflower mosaic virus 35S either at N- or C-terminal fusions are hardly detectable. We could detect GFP only by expression using a far stronger promoter, CvMV. In addition, previously we showed that the levels of DWF4 transcription are directly responsible for the hypocotyl length (Choe et al., 2001). Knock-out mutant, wild type, and 35S:DWF4 gradually, in this order increased the hypocotyl length proportional to the level of DWF4 expression. More importantly, we showed that the DWF4:GUS-expressing tissues are positively correlated with the organs that possess a higher level of bioactive BRs. The organs such as collets and roots express DWF4 and thus contain elevated amounts of bioactive BRs. Thus it is conceivable that Arabidopsis meets the requirement of BR-dependent growth through precise regulation of the DWF4 gene expression. Identification and functional characterization of the trans-acting factors that act on the DWF4 promoter should reveal the mechanism as to how the environmental and developmental signals regulate DWF4 gene expression.

Plant Materials, Growth Conditions, and Transgenic Plants Harboring DWF4:GUS Constructs

Two different reporter constructs were made. DNA fragments of 1,132 bp and 180 bp upstream from AUG translation start codon, respectively, were placed in front of the GUS (UidA) reporter gene in the pBI101 binary vector. The constructs were introduced into Arabidopsis (Arabidopsis thaliana) Wassilewskija-2 (Ws-2) wild type using conventional spray transformation protocols (Choe et al., 2001). T1 seeds were screened for transformants that are resistant to kanamycin (50 μg/mL) dissolved in agar-solidified medium. Forty independent transformants were isolated for each construct and grown to maturity. The T2 seeds harboring the constructs were stained using 5-bromo-4-chloro-3-indolyl-β-d-GlcUA according to the method described in Jefferson et al. (1987).

Through genetic crosses, the 1,132 bp promoter-containing DWF4:GUS construct was introduced into various genotypes including cpd-388, dwf12-1D, bzr1-D, and bri1-5. Putative DWF4:GUS lines in the different genetic backgrounds were selected from F₂ populations that show both morphological phenotypes and kanamycin resistance. Doubly homozygous lines for each mutation and the DWF4:GUS transgene were further confirmed at the F₃ generation. The doubly homozygous lines were plated on agar-solidified medium supplemented with mock or 0.1 μm BL and grown for 8 (light) and 3 d (dark) before GUS staining.

Feeding Tests, RT-PCR, and RNA Gel-Blot Analysis

Conditions for plant growth and feeding tests were described previously in Kwon et al. (2005). Briefly, the Arabidopsis ecotype Ws-2 seeds were surface sterilized with solution containing 20% (v/v) Clorox and 0.1% (w/v) SDS for 10 min and washed five times with sterile distilled water. The seeds were germinated in Murashige and Skoog liquid media (1× Murashige and Skoog basal salt mixture, 1% (w/v) Suc, pH 5.8 with KOH) for 7 d in a growth chamber (16-h light/8-h dark at 23°C) with continuous shaking at 100 rpm. The seedlings were pretreated with 100 μm of CHX for 2 h and then the culture media were supplemented with 1 μm epi-BL and 1 μm Brz for further 6 h of incubation with continuous shaking at 100 rpm. The harvested seedlings were quickly frozen in liquid nitrogen before proceeding to total RNA isolation. Total RNA was isolated using TRIzol reagent (Sigma) and further purified again using phenol-chloroform extraction. The purified total RNA (2 μg) was used for first-strand cDNA synthesis using the Superscript RNaseH⁻ reverse transcriptase (Invitrogen). The conditions for PCR amplification were an initial denaturation step for 5 min at 96°C, amplification by 25 cycles of 15 s at 94°C, 30 s at 50°C, and 1 min at 72°C, followed by final extension for 5 min at 72°C. Oligonucleotide sequences used for RT-PCR analysis are as follows; for DWF4, D4RTF, 5′-TTCTTGGTGAAACCATCGGTTATCTTAAA-3′, D4RTR, 5′-TATGATAAGCAGTTCCTGGTAGATTT-3′; and for CPD, CPDRT-F1, 5′-GTTTAATCTTGATTCTTGGTCTTCT-3′, CPDRT-R1, 5′-ATTTTATAACCTTTGATCTCAACAT-3′. The amplified PCR products were separated on a 1.2% (w/v) agarose gel, transferred to a nylon membrane, and probed with ³²P-labeled DWF4 and CPD genes. For staining with GUS after CHX treatment, 8-d-old, light-grown DWF4:GUS seedlings were pretreated for 2 h with 100 μm CHX and then the culture media were supplemented with 1 μm epi-BL and 1 μm Brz. After 24 h incubation, seedlings were stained overnight using 5-bromo-4-chloro-3-indolyl-β-d-GlcUA according to the method described in Jefferson et al. (1987).

To determine the DWF4 transcript level, 15 μg of total RNA isolated from various tissue types of mature Arabidopsis, 5-d-old dark-grown seedlings, and 3-week-old callus in standard callus-inducing media were fractionated in denaturing gel and subjected to the RNA gel-blot analysis according to the standard protocol (Sambrook et al., 1989).

Endogenous BR Analysis

For analysis of endogenous levels of BRs in different tissues, Arabidopsis plants were grown on soil (SUNSHINE MIX no. 5, SunGro) for 5 weeks under a long-day condition. Mature siliques, the joint tissues of shoot and root, shoot tips including developing floral organs, stems, and rosette leaves of 5-week-old Ws-2 wild-type plants were separated to collect up to 30 g of fresh tissues. Similarly, different tissues of bri1-5 were collected from population of 7-week-old plants. To obtain shoot and root tissues of seedlings, Ws-2 seeds were plated in a row on agar-solidified medium, and the plates were placed in a vertical position to have the roots to grow on the surface of agar media. The root and shoot tissues were separated with a razor blade before freezing in liquid nitrogen and further processing for steroid purification.

BR purification and quantification were carried out according to the method described by Noguchi et al. (1999). Gas chromatography/mass spectrometry analysis was done on a mass spectrometer (JMS-AM SUN200, JEOL) connected to a gas chromatograph (6890A, Agilent Technologies) with a capillary column DB-5 (0.25 mm × 15 m, 0.25 μ film thickness, J&W Scientific).

Subcellular Localization of DWF4

The DWF4 cDNA was obtained by RT-PCR, and cloned as an N-terminal GFP fusion in a plant expression vector containing CvMV promoter (Verdaguer et al., 1998) and nopaline synthase terminator. The CvMV:GFP-DWF4 construct was cotransfected with the 35S:BiP-RFP that is an ER marker protein (Koizumi, 1996) into Arabidopsis mesophyll protoplasts by polyethylene glycol-mediated transformation as described by Hwang and Sheen (2001). GFP and red fluorescent protein (RFP) fluorescences were observed using a confocal laser scanning microscope (Carl Zeiss LSM 510). The filter sets, BP505 to 530 (excitation 488 nm, emission 505–530 nm), BP560 to 615 (excitation 543 nm, emission 560–615 nm), and LP650 (excitation 488 nm, emission 650 nm) for GFP, RFP, and the chlorophyll autofluorescence, respectively, were used.

Sequence data from this article can be found in the GenBank/EMBL data libraries under accession number U12639.