• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of plntphysLink to Publisher's site
Plant Physiol. Mar 2007; 143(3): 1152–1162.
PMCID: PMC1820926

AGF1, an AT-Hook Protein, Is Necessary for the Negative Feedback of AtGA3ox1 Encoding GA 3-Oxidase1,[W]


Negative feedback is a fundamental mechanism of organisms to maintain the internal environment within tolerable limits. Gibberellins (GAs) are essential regulators of many aspects of plant development, including seed germination, stem elongation, and flowering. GA biosynthesis is regulated by the feedback mechanism in plants. GA 3-oxidase (GA3ox) catalyzes the final step of the biosynthetic pathway to produce the physiologically active GAs. Here, we found that only the AtGA3ox1 among the AtGA3ox family of Arabidopsis (Arabidopsis thaliana) is under the regulation of GA-negative feedback. We have identified a cis-acting sequence responsible for the GA-negative feedback of AtGA3ox1 using transgenic plants. Furthermore, we have identified an AT-hook protein, AGF1 (for the AT-hook protein of GA feedback regulation), as a DNA-binding protein for the cis-acting sequence of GA-negative feedback. The mutation in the cis-acting sequence abolished both GA-negative feedback and AGF1 binding. In addition, constitutive expression of AGF1 affected GA-negative feedback in Arabidopsis. Our results suggest that AGF1 plays a role in the homeostasis of GAs through binding to the cis-acting sequence of the GA-negative feedback of AtGA3ox1.

Homeostasis is one of the fundamental characteristics of organisms to regulate their internal environment so as to maintain stable conditions by multiple dynamic equilibrium adjustments. Negative feedback is a central mechanism of homeostasis, responding in a way that reverses the direction of changes. In this self-regulating system, the output response reduces the initial stimulus to keep fluctuations within limits acceptable to the function of organisms.

GAs, which are tetracyclic diterpenoid growth factors, are essential regulators of many aspects of plant development, including seed germination, stem elongation, flower induction, and anther development. A series of genes that encode the enzymes involved in the GA biosynthetic pathway has been cloned from various species, and an almost complete picture of GA biosynthesis has been revealed (for review, see Hedden and Phillips, 2000; Olszewski et al., 2002). Both endogenous developmental programs and environmental stimuli affect the expression of these enzymes. Therefore, elucidating the transcriptional regulation of GA biosynthetic enzymes is crucial to identify the molecular mechanisms involved in plant development and to understand how these mechanisms help plants adapt to changes in their environment. A factor that affects the transcription of GA biosynthetic enzymes is GA itself. In GA-deficient Arabidopsis (Arabidopsis thaliana) mutants (e.g. ga1–3), the expression of AtGA3ox1 (GA4) encoding GA 3-oxidase (GA3ox; formerly GA 3β-hydroxylase) and GA5 encoding GA 20-oxidase (GA20ox) is higher than that in the wild type, and this increased expression can be reduced by GA application (Chiang et al., 1995; Phillips et al., 1995; Xu et al., 1995). The mechanisms of negative feedback seem to contribute to the homeostasis of GA levels. Although the GA feedback regulation has been shown to depend on GA signaling components, including DELLA regulators, SPINDLY, PHOTOPERIOD-RESPONSIVE 1 (for review, see Fleet and Sun, 2005), and a GA receptor (Ueguchi-Tanaka et al., 2005), its molecular mechanisms are still largely unknown. Biosynthesis and signaling are important areas of phytohormone research; however, they have been studied independently thus far. The identification of the transcription factors that regulate the GA-negative feedback should provide new insights into the link between the control of endogenous amounts of GAs and intracellular GA signaling.

GA3ox catalyzes the final step of the biosynthetic pathway to produce the physiologically active GAs. Therefore, the expression of GA3ox genes is likely to play a key regulatory role in controlling the appropriate levels of active GAs during plant growth. Because GA3ox is encoded by multiple genes in several plants (Hedden and Phillips, 2000), it is also important to analyze the expression pattern of all members of the gene family to know where active GAs are produced in plants and which genes of the family play a central role in GA-negative feedback.

The AT-hook motif is a positively charged stretch of 13 amino acids containing the invariant core peptide sequence RGRP and is usually flanked by basic residues (Reeves and Nissen, 1990; Aravind and Landsman, 1998). It binds to DNA through a minor groove with the optimal DNA-binding site centered at the sequence AA(T/A)T (Maher and Nathans, 1996; Aravind and Landsman, 1998). The AT-hook was first identified in the high mobility group of non-histone chromosomal proteins, called HMGA (formerly HMG-I/Y). HMGA binding to DNA induces changes in DNA structure that lead to the formation of multinucleoprotein complexes called “enhanceosomes” in the regulatory promoter elements of the gene (Thanos and Maniatis, 1995). Since its discovery, the AT-hook motif has been found in either single or multiple copies in a large number of DNA-binding proteins, many of which are transcription factors or components of chromatin-remodeling complexes from a wide range of organisms (Aravind and Landsman, 1998). In plants, the AT-hook motif was found in nuclear matrix proteins that bind to special regions of chromosomal DNA called matrix attachment regions (Morisawa et al., 2000; Fujimoto et al., 2004).

In this study, we found that only the AtGA3ox1 among the AtGA3ox family is under regulation of GA-negative feedback. We have identified a cis-acting sequence, designated GNFEI, responsible for GA-negative feedback of AtGA3ox1. Furthermore, we have isolated an AT-hook protein, designated AGF1, as a DNA-binding protein for GNFEI. Constitutive expression of AGF1 enhanced up-regulation of AtGA3ox1 in response to decrease of GA, and the mutation in GNFEI abolished both GA-negative feedback and AGF1 binding. These results provide evidence that GNFEI and AGF1 are novel cis-trans regulatory factors involved in the homeostasis of GAs.


Expression Patterns of AtGA3ox Genes

The Arabidopsis GA3ox gene family consists of four members. At1g15550 (AtGA3ox1) and At1g80340 (AtGA3ox2) have been demonstrated to encode active GA3ox (Williams et al., 1998; Yamaguchi et al., 1998). Although the enzymatic activity of At4g21690 (AtGA3ox3) and At1g80330 (AtGA3ox4) was not examined, these genes were regarded as members of the GA3ox family because of their high similarities to AtGA3ox1 (see Supplemental Fig. S1). Recent work suggested that AtGA3ox1 and 2 are responsible for vegetative growth; however, they are dispensable for reproductive development (Mitchum et al., 2006). To obtain a further insight into the role of each gene of this family in GA homeostasis, we investigated which of the members of AtGA3ox family are regulated by feedback mechanism by reverse transcription (RT)-PCR and DNA gel-blot hybridization. First, we tried to determine where each AtGA3ox is expressed. Figure 1A shows that AtGA3ox1 was expressed in seedlings, leaves, stems, floral tips, and flowers. The mRNA for AtGA3ox2 was highly expressed in seedlings but was barely detectable in other tissues, as reported previously (Yamaguchi et al., 1998). AtGA3ox3 mRNA was detected in floral tips and slightly in flowers. AtGA3ox4 mRNA was detected only in flowers among examined organs (Mitchum et al., 2006). These results indicated that AtGA3ox1 is a major supplier of GA3ox enzymatic activity in any developmental stage. Based on the expression pattern, we next examined in seedlings and flowers the effects of applying GA or an inhibitor for GA biosynthesis, uniconazole P. The expression of AtGA3ox1 was regulated in a GA-negative feedback manner in both seedlings and flowers (Fig. 1B), as reported previously (Chiang et al., 1995). However, the expression of the other three AtGA3ox genes was not affected by the treatment with GA or uniconazole P (Fig. 1B). These results suggested that AtGA3ox1 plays a central role in the homeostasis of GA levels in plants. Thus, we focused on AtGA3ox1 to identify the molecular mechanisms involved in the transcriptional regulation of GA-negative feedback.

Figure 1.
Comparison of the expression patterns of AtGA3ox genes. A, RT-PCR analysis of AtGA3ox genes. RNAs were isolated from seedlings of 5-d-old Arabidopsis plants, leaves and first internode of 21-d-old plants, and floral tips and flowers of 35-d-old plants. ...

Identification of cis-Regions for the GA-Negative Feedback of AtGA3ox1

To define the cis-acting elements responsible for the GA-negative feedback of AtGA3ox1, we generated transgenic Arabidopsis plants carrying a series of 5′ deletions of the AtGA3ox1 promoter fused to a β-glucuronidase (GUS) reporter gene. We examined the effect of applying an inhibitor for GA biosynthesis, uniconazole P, or GA to the transgenic plants. Uniconazole P increased GUS activity, whereas GA decreased it for the Δ-2008 construct (Fig. 2A), showing that the fusion gene was under the control of GA-negative feedback. Although deletion of 992 bp from −2,008 to −1,017 (Δ-1016) resulted in a decrease of GUS activity, the feedback regulation was clearly preserved. Deletion of 208 bp from −1,016 to −809 (Δ-808) resulted in loss of GA-negative feedback. Conversely, the expression of Δ-808 was induced by GA. Therefore, there should be a cis-element(s) responsible for GA-negative feedback between −1,016 and −809 in the AtGA3ox1 promoter.

Figure 2.
Identification of cis-regions for GA-negative feedback of AtGA3ox1. A, Fluorometric GUS assay for GA-negative feedback in transgenic Arabidopsis plants that carried a series of deletions of the AtGA3ox1 promoter. Schematic diagrams of AtGA3ox1 promoter ...

To examine the function of the 208-bp sequence between −1,016 and −809 in the feedback regulation of GA, we carried out a gain-of-function experiment. Five tandem copies of the 208-bp DNA fragment of the AtGA3ox1 promoter were cloned upstream of a minimal promoter-GUS (208 × 5). The 208-bp DNA fragment endowed the reporter with a remarkable property of feedback regulation (Fig. 2B), suggesting that the sequences are sufficient for GA- negative feedback.

To further define cis-elements for the GA-negative feedback of the AtGA3ox1 promoter, 5′ deletion analysis of the 208-bp sequence between −1,016 and −809 was performed. Deletion of 56 bp from −1,016 to −961 (Δ-960) resulted in loss of the feedback regulation by GA (Fig. 2C). Consistently, Δ-918 and Δ-857 were not under the control of GA-negative feedback. The expression of Δ-857 was induced by GA; so was that of Δ-808. One possible explanation of this observation is that a GA-responsive element might exist downstream of −808 in the AtGA3ox1 promoter and be suppressed by the sequence upstream of −857. Furthermore, addition of the 43-bp sequence from −1,003 to −961 into Δ-918 recovered the GA-negative feedback (Δ-961/-918; Fig. 2C). These results substantiate the importance of the 43-bp sequence in the feedback regulation of GA. In summary, our results show that the 208-bp sequence between −1,016 and −809 is sufficient for GA-negative feedback and that the 43-bp sequence from −1,003 to −961 is indispensable. We named the 43-bp sequence GNFEI (for GA-negative feedback element I) and focused on the trans-acting factor that binds to it.

Identification of AT-Hook Proteins as the GNFEI-Binding Protein

To identify transcription factors that bind directly to GNFEI, we performed a yeast one-hybrid screen using 3 × GNFEI as bait on the Arabidopsis cDNA library. Screening of 3 × 106 transformants yielded two clones. A sequencing and database search revealed that both clones contained an AT-hook motif. Our cDNA clones represented two distinct genes, At4g35390 and At4g55560. We isolated the cDNA clones of At4g35390 and At4g55560 and confirmed the binding of the proteins to GNFEI in a yeast one-hybrid assay. We named these proteins AGF1 (for AT-hook protein of GA feedback regulation; At4g35390) and AGF2 (At3g55560). AGF1 was chosen for further analysis. An AT-hook motif and a conserved domain PPC (for plants and prokaryotes conserved domain) are found in the central part of this family of proteins (Fig. 3, A and B). Unlike HMGA proteins that contain three or four repeats of the AT-hook motif, AGF1 has only a single AT-hook motif. AGF1 belongs to a gene family consisting of 29 members in Arabidopsis. The family members break into two major evolutionary clades in the phylogenetic analysis (Fig. 3C). A nuclear matrix protein AHL1 (for At-hook motif nuclear localized protein 1) of Arabidopsis (Fujimoto et al., 2004) belongs to a distinct clade from that containing AGF1.

Figure 3.
Comparison of amino acid sequences of AGF1 and related proteins. A, Amino acid sequence of AGF1. The AT-hook motif is boxed, and the following PPC domain is underlined. B, Alignment of the amino acid sequences of the AT-hook domain and the PPC domain ...

Binding of AGF1 to GNFEI

To confirm the specific binding of AGF1 to GNFEI, a gel retardation assay was performed. AGF1 protein prepared by in vitro translation formed a specific complex with a 32P-labeled GNFEI fragment (Fig. 4A). Six copies of the AAAT sequence were found in GNFEI (Fig. 4B). Because it has been shown that the AT-hook motif binds to an AT-rich sequence (Aravind and Landsman, 1998), the repeat of the AAAT sequence could be the target of AGF1. To determine the nucleotides that are important for binding to AGF1, we tested mutant variants of the GNFEI for AGF1 binding in gel-shift assays. Mutations in repeats 1 and 2 (Mt1) or in repeats 5 and 6 (Mt3) resulted in only a slight reduction of AGF1 binding; however, mutations in repeats 3 and 4 (Mt2) diminished the ability to interact with AGF1 (Fig. 4C). Furthermore, the formation of a complex of wild-type GNFEI with AGF1 was not inhibited in the presence of an excess amount of the mutated sequence Mt2 (data not shown). These results indicate that AAAT sequences in the central region of GNFEI (repeats 3 and 4) are important for AGF1 binding, whereas adjacent repeats are dispensable for the binding.

Figure 4.
Sequence-specific binding of AGF1. A, Gel retardation assay with in vitro-translated AGF1. Oligonucleotides containing GNFEI were used as a probe. The specific AGF1-DNA complexes are indicated by an arrowhead. +, Addition to the reaction mixtures; ...

We examined the transcriptional activity of AGF1 in yeasts. AGF1 itself, as well as a fusion protein of AGF1 with the GAL4 activation domain, activated the transcription of 3 × GNFEI-HIS3 in the yeast one-hybrid assay (Fig. 4D), indicating that AGF1 is a transcriptional activator, at least in yeasts.

AGF1 Binding Is Required for GA-Negative Feedback

To determine whether in vitro binding of AGF1 to GNFEI is consistent with in vivo GA-negative feedback, we constructed a mutant version of the Δ-1016 AtGA3ox1 promoter-GUS in which the mutation corresponding to Mt2 was introduced to eliminate the ability to bind with AGF1 (Fig. 5A). As expected, Mt2 mutation abolished the transcriptional regulation of the GA-negative feedback of the AtGA3ox1 promoter in the transgenic Arabidopsis plants (Fig. 5A). This loss-of-function experiment showed that the AAAT sequences in the central region of GNFEI (repeats 3 and 4) are necessary for both GA-negative feedback and AGF binding, suggesting that AGF1 is involved in the feedback regulation of AtGA3ox1 through binding to GNFEI.

Figure 5.
Effect of mutation in GNFEI on the GA-negative feedback of AtGA3ox1. A, Fluorometric GUS assay for GA-negative feedback in transgenic Arabidopsis plants carrying the Mt2 mutant version of Δ-1016. The white bars represent the GUS activity of plants ...

To examine whether Mt2 mutation affected the spatial expression pattern of AtGA3ox1, we compared the patterns of GUS staining in transgenic plants carrying the wild-type Δ-1016 construct with those of the Mt2 mutant version. Histochemical analysis indicated that the wild-type AtGA3ox1-GUS reporter gene was expressed in hypocotyls, cotyledons, petioles, and shoot apices (Fig. 5B, a and insert). However, the Mt2 mutant version was not expressed in the shoot apices (Fig. 5B, c and insert). These observations suggest that AGF binding to GNFEI participates in GA-negative feedback as well as in the control of expression in the shoot apices of AtGA3ox1.

We also examined whether the decrease of GA levels alters the pattern of the tissue-specific expression of AtGA3ox1 or simply elevates the levels of expression in those same tissues. The application of uniconazole P elevated the AtGA3ox1 mRNA levels in organ regions and tissues that normally express AtGA3ox1 (Fig. 5B, b). The expression pattern of the Mt2 mutant version was not affected by uniconazole P (Fig. 5B, d). Our results show that AtGA3ox1 was regulated in a tissue-specific manner and this regulation was maintained even when the endogenous GA levels decreased. The expression pattern of AtGA3ox1-GUS in normal conditions was independently reported (Mitchum et al., 2006). Although they observed strong staining in roots, we detected only weak GUS activities in roots. This discrepancy might be reflecting the promoter length of the GUS fusion genes; we used 1 kb of 5′ upstream region, while they used 3 kb. In other organs, both observations were similar.

Expression Pattern of AGF1

The expression pattern of the AGF1 gene was investigated by RT-PCR analysis, with total RNAs isolated from Arabidopsis plants. The mRNAs for AGF1 are expressed in seedlings, leaves, stems, floral tips, and flowers (Fig. 6A). We next investigated whether the transcription of the AGF1 gene is regulated by the GA levels. Figure 6B shows that the AGF1 mRNA levels were not affected by the treatment with either GA or uniconazole P. This result suggests that the function of AGF1 is regulated at the posttranscriptional and/or posttranslational levels.

Figure 6.
GAs do not affect the intracellular localization of AGF1. A, RT-PCR analysis of AGF1. Total RNAs isolated from the indicated organs were used for RT-PCR analysis. After amplification, the products were detected by DNA gel-blot hybridization. β ...

Intracellular Localization of AGF1

RSG (for REPRESSION OF SHOOT GROWTH) is a tobacco (Nicotiana tabacum) transcriptional activator with a basic Leu zipper domain that is involved in the regulation of endogenous amounts of GAs (Fukazawa et al., 2000). The intracellular localization of RSG is regulated by the endogenous amounts of GAs (Ishida et al., 2004). We examined the intracellular localization of AGF1 using transgenic Arabidopsis plants in which the fusion gene of AGF1 and green fluorescent protein (GFP) was expressed under the control of the 35S promoter of the Cauliflower mosaic virus. Consistent with the observation that AGF1 is a DNA-binding protein, AGF1-GFP was localized exclusively in the nucleus in the epidermal cells of petioles in transgenic plants (Fig. 6C). Next, we examined the effects of applying GA or an inhibitor for GA biosynthesis, uniconazole P, to transgenic plants in which AGF1-GFP was expressed. Figure 6C shows that the subcellular localization of AGF1-GFP was not affected by the treatment with either GA or uniconazole P. These results suggest that the function of AGF1 is regulated by mechanisms other than the control of intracellular localization.

Function of AGF1 in GA-Negative Feedback of AtGA3ox1

To confirm the function of AGF1 in the GA-negative feedback, we generated transgenic Arabidopsis constitutively expressing AGF1. Because the expression level of AtGA3ox1 of the AGF1 constitutive expressor in 5-d-old seedlings was comparable to that of the wild type, AGF1 did not seem to be a general enhancer of AtGA3ox1. But the feedback up-regulation of AtGA3ox1 by uniconazole P application was enhanced in AGF1 constitutive expressor (Fig. 7A). If AGF1 plays a role as a transcriptional activator in response to decrease of GA levels, constitutive expression of AGF1 should suppress the transcriptional repression of AtGA3ox1 by GA. GA application after uniconazole P treatment immediately repressed the expression of AtGA3ox1 in wild type to 30% at 30 min and 10% at 60 min. However, this feedback repression was attenuated in AGF1 constitutive expressor, i.e. the expression level of AtGA3ox1 was maintained at 80% at 30 min and 30% at 60 min (Fig. 7B). Because AGF1 constitutive expression conferred both hyper up-regulation of AtGA3ox1 by reduction of GA levels and resistance to down-regulation of AtGA3ox1 by excess GA, AGF1 may be a transcriptional activator specific to GA-negative feedback.

Figure 7.
Constitutive expression of AGF1 altered the GA-negative feedback of AtGA3ox1. A, A time-course study of feedback activation of AtGA3ox1 by uniconazole P (Uni). RNAs were isolated from both wild-type plants and AGF1 constitutive expressor (AGF1ce) harvested ...


Homeostasis is the maintenance of constant conditions by means of automatic mechanisms that counteract influences tending toward disequilibrium. Any homeostatic system requires three functional components: a receptor, a controller, and an effector. In the transcriptional regulation of GA-negative feedback, the controller processes information from a GA receptor and directs the transcription of the effectors, the genes for GA3ox and GA20ox. Identification of transcription factors for the feedback regulation is the key to elucidating the molecular mechanism for GA homeostasis.

In this study, we have identified a cis-acting sequence and a DNA-binding protein (AGF1) responsible for the GA-negative feedback of AtGA3ox1. AGF1 contains an AT-hook motif that binds to the minor groove of AT-rich DNA sequences. AT-hook motifs are frequently associated with other known functional domains seen in chromatin proteins and DNA-binding proteins. In addition to the AT-hook, Drosophila melanogaster apterous as well as mammalian LH2 and Barx1 also contain homeodomains; human RFX5 protein also contains the DNA-binding RFX box; human ESE-1 protein also contains an ETS domain; and Saccharomyces cerevisiae SWI5, Drosophila castor, and mouse CTCF1 also contain zinc fingers (Aravind and Landsman, 1998). In these cases, the AT-hook might provide a secondary association with a neighboring AT-rich sequence, possibly stabilizing an interaction primarily mediated by another DNA-binding domain. Although no other known DNA-binding motif was found in AGF1, it contains a PPC domain. The PPC domain is evolutionarily conserved in bacteria, archaea, and plants, but not in animals. A function of the PPC domain is not apparent at present; however, it might be involved in molecular and cellular functions, including DNA binding, interaction with other proteins, and subcellular localization, because the hydrophobic region of PPC in AHL1 of Arabidopsis was indispensable for nuclear localization (Fujimoto et al., 2004). Twenty-nine AGF1-related genes were found in the Arabidopsis genome and 15 of them exhibited higher similarity (Fig. 3). Thus, the absence of obvious phenotypes for a T-DNA insertion mutant of AGF1 under normal growth conditions (data not shown) is likely attributable to the functional redundancy among the members of the family. Overexpression of AGF1 or AGF1-GFP did not promote the expression of AtGA3ox1 (data not shown), suggesting that AGF1 is necessary but not sufficient for up-regulation of AtGA3ox1.

The genes of GA20ox as well as those of GA3ox are regulated by feedback mechanisms in several plants, including Arabidopsis and tobacco (Hedden and Phillips, 2000). Meier et al. (2001) showed that the cis-element(s) responsible for GA-negative feedback is located within 500 bp upstream of the transcription start site of GA5 encoding GA20ox. However, a DNA sequence similar to GNFEI of AtGA3ox1 was not found in the GA5 promoter. In this context, it is interesting to note that the feedback regulation of the GA3ox gene was not affected in the transgenic tobacco plants in which the function of RSG was repressed, although that of the GA20ox gene was impaired in the same plants (Ishida et al., 2004). These results suggest that the feedback regulation of the GA20ox gene and the GA3ox gene is based on different mechanisms, at least in part. Furthermore, the KNOX homeodomain protein NTH15 (for N. tabacum homeobox 15) repressed the expression of Ntc12 encoding GA20ox but not Nty2 encoding GA3ox in tobacco plants (Sakamoto et al., 2001). On the other hand, the overexpression of FUS3 (for FUSCA3), a member of the plant-specific transcription factor B3 family, repressed AtGA3ox1 but did not affect the expression of AtGA20ox1 in Arabidopsis (Gazzarrini et al., 2004). GA 20- and GA 3-oxidases catalyze the penultimate and final steps, respectively, in the formation of bioactive GAs in the cytoplasm; however, their transcripts accumulate differentially in tissues (Ueguchi-Tanaka et al., 1998; Itoh et al., 1999). Such independent transcriptional regulations of GA biosynthetic enzymes might provide elaborate mechanisms to regulate the endogenous level of GAs in response to internal and external stimuli.

Another finding is that AtGA3ox2 and AtGA3ox4, which are not under regulation of GA feedback, are highly expressed in seedlings and flowers, respectively (Fig. 1A). The feedback regulation contributes to the homeostasis of GA levels; however, this mechanism could prevent the accumulation of GAs, which may be essential for seed germination and flowering. Therefore, the expression of AtGA3ox2 and AtGA3ox4 in seedlings and flowers may be crucial to circumvent this potential problem and to produce sufficient bioactive GAs for seed germination and flowering.

Although GA biosynthesis is restricted to specific regions, including actively growing and elongating tissues (Smith, 1992), AGF1 was expressed ubiquitously in the plant organs that we examined (Fig. 6A). Although it is possible that the expression of AGF1 is regulated in a cell type-specific manner within each organ, this apparent inconsistency suggests an involvement of posttranscriptional and/or posttranslational modifications of the transcription factor. Unchanged expression of AtGA3ox1 in AGF1 constitutive expressor in normal growth condition also suggests the existence of a functional regulation of AGF1 in response to decrease of GA levels. We show that the intracellular localization of AGF1 was not affected by the GA levels (Fig. 6C). One possible mechanism for the functional regulation of AGF1 is the interaction of AGF1 with accessory proteins, including other transcriptional activators, repressors, general transcription factors, coactivators, or chaperones. From the viewpoint of GA signaling, it is important to consider the relationship between AGF1 and other GA signaling components. Although recent progresses have revealed that the degradation of DELLA proteins induced by GAs is critical in GA signaling, the targets of DELLA proteins remain unknown. The expression of AtGA3ox1 was repressed in the rga mutant (Silverstone et al., 1998), suggesting that GA-negative feedback of AtGA3ox1 is regulated by DELLA proteins. It is possible that DELLA proteins directly regulate AGF1. However, in our preliminary study, AGF1 could not interact with RGA at least in yeasts (A. Matsushita, unpublished data). Therefore, additional factors might be required for DELLA proteins to regulate AGF1 and GA-negative feedback.

Alternatively, covalent modification, including acetylation, methylation, and phosphorylation, could be involved in the regulation of AGF1. The HMGA protein that contained AT-hook motifs plays a role in the virus-induced transcriptional activation of the human IFN-β gene (Thanos and Maniatis, 1995). HMGA organizes the enhanceosome into a structure that optimally interacts with chromatin-modifying activities and general transcription factors. Acetylation of HMGA by p300/CBP-associated factors/GCN5 at Lys-71 potentiates the transcription by stabilizing the enhanceosome, whereas acetylation of HMGA by the CREB-binding protein (CBP) at Lys-65 destabilizes the enhanceosome (Munshi et al., 2001). Thus, the ordered and highly controlled acetylation of HMGA by two distinct histone acetyltransferase coactivators coordinates the IFN-β transcriptional switch by instructing either enhanceosome assembly or disassembly. Decreases in the DNA-binding affinity of animal HMGA proteins are brought about by phosphorylation by a number of kinases, including Casein kinase II (Wang et al., 1995; Schwanbeck et al., 2001), Cdc2 kinase (Nissen and Reeves, 1995), mitogen-activated protein kinase (Schwanbeck and Wiśniewski, 1997), and protein kinase C (Xiao et al., 2000). These covalent modifications might also decrease protein-protein interactions or create novel interactions.

BZR1 (for BRASSINAZOLE-RESISTANT 1) of Arabidopsis is a transcriptional repressor that is involved in both brassinosteroid (BR)-mediated growth promotion and the feedback regulation of BR biosynthesis (Wang et al., 2002). When the BR level is low, BZR1 is phosphorylated and degraded by the proteasome (He et al., 2002). BR promotes the dephosphorylation of BZR1, which prevents the degradation of BZR1 and increases its accumulation in the nucleus. BZR1 then binds to the promoters and inhibits the expression of CPD (for constitutive photomorphogenesis and dwarfism), a BR biosynthetic gene (He et al., 2005). AGF1 could be negatively regulated by GAs through phosphorylation and could be activated by dephosphorylation in response to a decrease of GA amounts in plants.

We found that Ser/Thr kinase activity is required for the apparent cytoplasmic localization of RSG that might play a role in the feedback regulation of the GA20ox gene (Ishida et al., 2004). The kinases promote the cytoplasmic migration of RSG by enhancing the association of RSG with 14-3-3 proteins downstream of GA signaling (S. Ishida, unpublished data). An attractive possibility is that identical kinases are involved in the phosphorylation of both RSG and AGF1 to quench their activities in response to GAs. Further investigation of how the function of AGF1 is controlled by GAs will help to reveal the molecular mechanisms for the fine regulation of GA-negative feedback and provide clues to understand the interaction between the control of endogenous amounts of GAs and GA signaling.


Yeast One-Hybrid Screen

To create the three tandem copies of GNFEI of the AtGA3ox1 for a one-hybrid screen, the PCR-amplified fragment (−1,003 to −961) with XbaI (5′ end) and SpeI (3′ end) was cloned into XbaI of pBluescript II. The fragment containing three tandem copies of GNFEI was excised by EcoRI and XhoI from the pBluescript II and cloned into the HIS3 reporter plasmid, pHISi (CLONTECH). The reporter construct was integrated into the genome of the yeast strain YM4271. The cDNA library of Arabidopsis (Arabidopsis thaliana) seedlings was screened by transformation of the reporter strains on plates without His but containing 3-aminotriazole (40 mm).

Plasmid Construction

The promoter of AtGA3ox1 (−2,016 to −1) was amplified by a PCR reaction from Arabidopsis genomic DNA with two primers, 5′-AAAGGATCCCTTGCTCTTTTTTAATTAGTTTTA-3′ and 5′-AAAGTCGACTCTTCCACTAAACAAAACTGGAAT-3′, generating a BamHI site and a SalI site at the 5′ and the 3′ ends, respectively. The amplified DNA fragment was digested with SalI/BamHI, cloned into a pUC18 vector, and sequenced. A series of 5′ deletion constructs was generated by PCR using this plasmid as a template. To make the internal deletion constructs, the PCR-amplified upstream fragments of the promoter with HindIII (5′ end) and SalI (3′ end) were cloned into HindIII and SalI of the 5′ deletion constructs. To make the five tandem copies of a 208-bp fragment of the AtGA3ox1 promoter (−1,016 to −808), the PCR-amplified fragment with XbaI (5′ end) and SpeI (3′ end) was cloned into XbaI of pUC18. Five tandem copies of the 208-bp DNA fragment of the AtGA3ox1 promoter were cloned upstream of the parB minimal promoter. The 5′ end of the parB minimal promoter was 4 bp upstream of the TATA box of parB (Takahashi et al., 1995). The integrity of all constructs was confirmed by DNA sequencing. These AtGA3ox1 promoter constructs were subcloned into transformation vector pBI101 to generate GUS fusion genes. For generating AGF1 constitutive expressor and GFP fusion constructs, AGF1 cDNA was amplified by specific primer pairs and cloned into pJ4 and pJ4-GFP vector, respectively (Fukazawa et al., 2000; Igarashi et al., 2001). These AGF1 constructs were subcloned into transformation vector pBI101.

Plant Material and Transformation

All Arabidopsis lines used were in the Columbia-0 background. Columbia-0 was used as the wild type. Seeds were sterilized and were sown on media containing 0.3% phytagel (Sigma) and half-strength Murashige and Skoog salts, and 1% Suc with or without antibiotic. The plates were incubated in the dark at 4°C for 3 d and subsequently transferred to a continuous light or long-day condition (18 h of light/6 h of dark) at 22°C. For hormonal treatments, 5-d-old seedlings were transferred to the plate containing 10 mg/L GA3 or 1 mg/L uniconazole P (Wako) and grown for 1 week. Transgenic Arabidopsis plants were produced by the floral-dip method (Clough and Bent, 1998) with Agrobacterium tumefaciens strain GV3101. Transformed plants were selected on plates containing 50 μg/mL kanamycin. Transgenic plants identified in this generation were classified as T1 plants.

GUS Enzyme Assay

Kanamycin-resistant transformants were histochemically stained to detect GUS activity by immersing seedlings in a staining solution (a 50 mm sodium phosphate buffer, pH 7.5, with 0.1% Triton X-100 and 1 mm 5-bromo-4-chloro-3-indolyl 6-β-glucuronide) overnight at 37°C. After staining, samples were immersed in a fixing solution (5% formaldehyde, 5% potassium acetate, 20% ethanol) followed by dechlorophylation in 70% ethanol.


For RT-PCR studies, total RNAs were converted into cDNAs using SuperScript III (Invitrogen) and oligo(dT)12–18 (Amersham Bioscience). PCR was performed with cDNA derived from 0.1 μg of total RNA with Extaq (Takara). The specific primer sequences were 5′-GCCGGATCCAATTAAAAAAGAGCAAGATGCCTGCTAT-3′ and 5′-AAAGGTACCTGTTCCTCGTACTCTTCAACGATATCG-3′ for AtGA3ox1, 5′-GCCGGGATCCCACATAAGCCTTTTAGCATGAGTTCAAC-3′ and 5′-GCCGGTACCTGTTCTTCATACTCTTCAATAATTTCA-3′ for AtGA3ox2, 5′-GCCGGATCCATCCTATAATTATAATCATGAGCTCTGT-3′ and 5′-AAAGGTACCGCCTTCTGATACTCTTCCATCACATTG-3′ for AtGA3ox3, 5′-GCCGGATCCTCATTAGTTCACAAGTCATGCCTTCACT-3′ and 5′-GCAGGTACCTCGTCCACGTATTCTTGGATTATACCG-3′ for AtGA3ox4, and 5′-GAAGTCGACCGTGAGATTCTTCACATCCAGGGTGGTC-3′ and 5′-CGCGGATCCCATAGTAGCAGAAATCAAGTGGTTCAAA-3′ for β-tubulin. The PCR products were separated by size on a 1% (w/v) agarose gel, blotted onto Biodyne B (Pall), and hybridized with DNA probes. The chemiluminescence signal was detected with an imaging system (LAS1000 plus; Fujifilm).

Gel Mobility Shift Assay

AGF1 protein was prepared in vitro using rabbit reticulocyte lysate as described previously (Igarashi et al., 2001). The oligonucleotide sequences of the GNFEI fragment used for the gel mobility shift assays were 5′-GAGTTTTGTATGTTCAAATAAATATTTATTTATTTTTATATATTT-3′ and 5′-GAGAAAATATATAAAAATAAATAAATATTTATTTGAACATACAAAA. The oligonucleotides were annealed and then labeled with [α-32P]dCTP and the Klenow fragment of DNA polymerase I. Binding mixtures contained 100 fmol of a labeled probe, in vitro-translated AGF1, and 2 μg of poly(dI-dC). Cold dsDNA fragments of GNFEI were used at 100-fold molar excess as DNA competitor. The binding buffer consisted of 25 mm HEPES-KOH, pH 7.9, 10 mm MgCl2, 50 mm KCl, 0.5 mm EDTA, 10% glycerol, and 0.5 mm dithiothreitol. Reactions were incubated at room temperature for 15 min and loaded onto 8% polyacrylamide gels containing 44.5 mm Tris, 44.5 mm borate, and 1 mm EDTA.

Fluorescent Microscopy

Epidermal cells of petioles of transgenic Arabidopsis seedlings were examined using an epifluorescence microscope (Nikon eclipse 80i) equipped with a CCD camera (Nikon DXM1200C).

Supplemental Data

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

  • Supplemental Figure S1. Alignment of AtGA3ox proteins.

Supplementary Material

[Supplemental Data]


We thank Sachiko Matsunaga for the generation of transgenic plants, and Jutarou Fukazawa and Risa Kanamoto for providing technical advice and participating in helpful discussions.


1This work was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture (Japan) to Y.T.

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instruction for Authors (www.plantphysiol.org) is: Yohsuke Takahashi (pj.ca.u-amihsorih@sahakaty).

[W]The online version of this article contains Web-only data.



  • Aravind L, Landsman D (1998) AT-hook motifs identified in a wide variety of DNA-binding proteins. Nucleic Acids Res 26 4413–4421 [PMC free article] [PubMed]
  • Chiang H-H, Hwang I, Goodman HM (1995) Isolation of the Arabidopsis GA4 locus. Plant Cell 7 195–201 [PMC free article] [PubMed]
  • Clough SJ, Bent AF (1998) Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J 16 735–743 [PubMed]
  • Fleet CM, Sun T-p (2005) A DELLAcate balance: the role of gibberellin in plant morphogenesis. Curr Opin Plant Biol 8 77–85 [PubMed]
  • Fujimoto S, Matsunaga S, Yonemura M, Uchiyama S, Azuma T, Fukui K (2004) Identification of a novel plant MAR DNA binding protein localized on chromosomal surfaces. Plant Mol Biol 56 225–239 [PubMed]
  • Fukazawa J, Sakai T, Ishida S, Yamaguchi I, Kamiya Y, Takahashi Y (2000) REPRESSION OF SHOOT GROWTH, a bZIP transcriptional activator, regulates cell elongation by controlling the level of gibberellins. Plant Cell 12 901–915 [PMC free article] [PubMed]
  • Gazzarrini S, Tsuchiya Y, Lumba S, Okamoto M, McCourt P (2004) The transcription factor FUSCA3 controls developmental timing in Arabidopsis through the hormones gibberellin and abscisic acid. Dev Cell 7 373–385 [PubMed]
  • He JX, Gendron JM, Yang Y, Li J, Wang ZY (2002) The GSK3-like kinase BIN2 phosphorylates and destabilizes BZR1, a positive regulator of the brassinosteroid signaling pathway in Arabidopsis. Proc Natl Acad Sci USA 99 10185–10190 [PMC free article] [PubMed]
  • He JX, Gendron JM, Sun Y, Gampala SS, Gendron N, Sun CQ, Wang ZY (2005) BZR1 is a transcriptional repressor with dual roles in brassinosteroid homeostasis and growth responses. Science 307 1634–1638 [PMC free article] [PubMed]
  • Hedden P, Phillips AL (2000) Gibberellin metabolism: new insights revealed by the genes. Trends Plant Sci 5 523–530 [PubMed]
  • Igarashi D, Ishida S, Fukazawa J, Takahashi Y (2001) 14-3-3 proteins regulate intracellular localization of the bZIP transcriptional activator RSG. Plant Cell 13 2483–2497 [PMC free article] [PubMed]
  • Ishida S, Fukazawa J, Yuasa T, Takahashi Y (2004) Involvement of 14-3-3 signaling protein binding in the functional regulation of the transcriptional activator REPRESSION OF SHOOT GROWTH by gibberellins. Plant Cell 16 2641–2651 [PMC free article] [PubMed]
  • Itoh H, Ueguchi-Tanaka M, Kawaide H, Chen X, Kamiya Y, Matsuoka M (1999) The gene encoding tobacco gibberellin 3β-hydroxylase is expressed at the site of GA action during stem elongation and flower organ development. Plant J 20 15–24 [PubMed]
  • Maher JF, Nathans D (1996) Multivalent DNA-binding properties of the HMG-I proteins. Proc Natl Acad Sci USA 93 6716–6720 [PMC free article] [PubMed]
  • Meier C, Bouquin T, Nielsen ME, Raventos D, Mattsson O, Rocher A, Schomburg F, Amasino RM, Mundy J (2001) Gibberellin response mutants identified by luciferase imaging. Plant J 25 509–519 [PubMed]
  • Mitchum MG, Yamaguchi S, Hanada A, Kuwahara A, Yoshioka Y, Kato T, Tabata S, Kamiya Y, Sun T-p (2006) Distinct and overlapping roles of two gibberellin 3-oxidases in Arabidopsis development. Plant J 45 804–818 [PubMed]
  • Morisawa G, Han-Yama A, Moda I, Tamai A, Iwabuchi M, Meshi T (2000) AHM1, a novel type of nuclear matrix-localized, MAR binding protein with a single AT hook and a J domain-homologous region. Plant Cell 12 1903–1916 [PMC free article] [PubMed]
  • Munshi N, Agalioti T, Lomvardas S, Merika M, Chen G, Thanos D (2001) Coordination of a transcriptional switch by HMGI(Y) acetylation. Science 293 1133–1136 [PubMed]
  • Nissen MS, Reeves R (1995) Changes in superhelicity are introduced into closed circular DNA by binding of high mobility group protein I/Y. J Biol Chem 270 4355–4360 [PubMed]
  • Olszewski N, Sun T-p, Gubler F (2002) Gibberellin signaling: biosynthesis, catabolism, and response pathways. Plant Cell (Suppl) 14 S61–S80 [PMC free article] [PubMed]
  • Phillips AL, Ward DA, Uknes S, Appleford NEJ, Lange T, Huttly A, Gaskin P, Graebe JE, Hedden P (1995) Isolation and expression of three gibberellin 20-oxidase cDNA clones from Arabidopsis. Plant Physiol 108 1049–1057 [PMC free article] [PubMed]
  • Reeves R, Nissen MS (1990) The A.T-DNA-binding domain of mammalian high mobility group I chromosomal proteins. J Biol Chem 265 8573–8582 [PubMed]
  • Sakamoto T, Kamiya N, Ueguchi-Tanaka M, Iwahori S, Matsuoka M (2001) KNOX homeodomain protein directly suppresses the expression of a gibberellin biosynthetic gene in the tobacco shoot apical meristem. Genes Dev 15 581–590 [PMC free article] [PubMed]
  • Schwanbeck R, Gymnopoulos M, Petry I, Piekiełko AI, Szewczuk Z, Zechel K, Wiśniewski JR (2001) Consecutive steps of phosphorylation affect conformation and DNA binding of the Chironomus high mobility group A protein. J Biol Chem 276 26012–26021 [PubMed]
  • Schwanbeck R, Wiśniewski JR (1997) Cdc2 and mitogen-activated protein kinases modulate DNA binding properties of the putative transcriptional regulator Chironomus high mobility group protein I. J Biol Chem 272 27476–27483 [PubMed]
  • Silverstone AL, Ciampaglio CN, Sun T-p (1998) The Arabidopsis RGA gene encodes a transcriptional regulator repressing the gibberellin signal transduction pathway. Plant Cell 10 155–169 [PMC free article] [PubMed]
  • Smith VA (1992) Gibberellin A1 biosynthesis in Pisum sativum L. II. Biological and biochemical consequences of the le mutation. Plant Physiol 99 372–377 [PMC free article] [PubMed]
  • Takahashi Y, Sakai T, Ishida S, Nagata T (1995) Identification of auxin-responsive elements of parB and their expression in apices of shoot and root. Proc Natl Acad Sci USA 92 6359–6363 [PMC free article] [PubMed]
  • Thanos D, Maniatis T (1995) Virus induction of human IFNβ gene expression requires the assembly of an enhanceosome. Cell 83 1091–1100 [PubMed]
  • Ueguchi-Tanaka M, Ashikari M, Nakajima M, Itoh H, Katoh E, Kobayashi M, Chow TY, Hsing YI, Kitano H, Yamaguchi I, et al (2005) GIBBERELLIN INSENSITIVE DWARF1 encodes a soluble receptor for gibberellin. Nature 437 693–698 [PubMed]
  • Ueguchi-Tanaka M, Itoh H, Oyama N, Koshioka M, Matsuoka M (1998) Over-expression of a tobacco homeobox gene, NTH15, decreases the expression of a gibberellin biosynthetic gene encoding GA 20-oxidase. Plant J 15 391–400 [PubMed]
  • Wang DZ, Ray P, Boothby M (1995) Interleukin 4-inducible phosphorylation of HMG-I(Y) is inhibited by rapamycin. J Biol Chem 270 22924–22932 [PubMed]
  • Wang ZY, Nakano T, Gendron J, He J, Chen M, Vafeados D, Yang Y, Fujioka S, Yoshida S, Asami T, et al (2002) Nuclear-localized BZR1 mediates brassinosteroid-induced growth and feedback suppression of brassinosteroid biosynthesis. Dev Cell 2 505–513 [PubMed]
  • Williams J, Phillips AL, Gaskin P, Hedden P (1998) Function and substrate specificity of the gibberellin 3 β-hydroxylase encoded by the Arabidopsis GA4 gene. Plant Physiol 117 559–563 [PMC free article] [PubMed]
  • Xiao D-M, Pak JH, Wang X, Sato T, Huang FL, Chen H-C, Huang K-P (2000) Phosphorylation of HMG-I by protein kinase C attenuates its binding affinity to the promoter regions of protein kinase C γ and neurogranin/RC3 genes. J Neurochem 74 392–399 [PubMed]
  • Xu Y-L, Li L, Wu K, Peeters AJM, Gage DA, Zeevaart JAD (1995) The GA5 locus of Arabidopsis thaliana encodes a multifunctional gibberellin 20-oxidase: molecular cloning and functional expression. Proc Natl Acad Sci USA 92 6640–6644 [PMC free article] [PubMed]
  • Yamaguchi S, Smith MW, Brown RGS, Kamiya Y, Sun T-p (1998) Phytochrome regulation and differential expression of gibberellin 3 β-hydroxylase genes in germinating Arabidopsis seed. Plant Cell 10 2115–2126 [PMC free article] [PubMed]

Articles from Plant Physiology are provided here courtesy of American Society of Plant Biologists
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...