Brassinosteroid homeostasis via coordinate regulation of signaling and synthetic pathways

Abstract

A widely accepted regulatory mechanism in maintaining hormone homeostasis involves negative or positive feedback control of biosynthetic genes through signal transduction pathways triggered by hormones. For brassinosteroid (BR) homeostasis, the antagonistic relationship between signaling and biosynthetic pathways has been well characterized. We have identified a transcriptional regulator, RAV-Like1, which activates both a BR receptor gene (BRI1) and BR synthetic genes (D2, D11 and BRD1). RAVL1 possesses a B3 DNA binding domain that exhibits differential affinity for E-box elements in the promoters of BRI1, D2, D11 and BRD1. Semi-dwarfism and BR-insensitive phenotypes are exhibited by ravl1 mutants. Genetic studies have demonstrated that expression alteration of BRI1 and BR synthetic genes by RAVL1 results in changes in BR sensitivity. BZR1 is a negative regulator involved in BR feedback mechanisms. To examine the relationship between RAVL1 and BZR1, expression of the common target gene BRD1 was examined using a transient transcription assay. The suppression of BRD1 by BZR1 is epistatic to activation by RAVL1. More importantly, RAVL1 is not subject to BR feedback regulation. Taken together, this data indicates that RAVL1 is involved in maintaining the basal activity of BRI1 and BR synthetic genes, which ensures that the basal levels of the hormone are produced. This study elucidated the RAVL1-mediated basal activation system which, in cooperation with negative feedback mechanisms, maintains BR homeostasis in higher plants.

RAVL1 mutants exhibit either BR insensitive or deficient phenotypes, while RAVL1 overexpressors exhibit hyper-active phenotypes.¹ The ravl1 mutant exhibits semi-dwarfism with thick and dark green leaves. In the dark, mutants failed to undergo skotomorphogenesis. In contrast, overexpressors developed exaggerated angles at the lamina joints between the blades and sheaths of the leaves. Alterations in BR-sensitivity in mutants and overexpressors were confirmed by examining the growth rates of seminal roots and bend angles of lamina joints in response to exogenous BR.

The mRNA levels of both BRI1 (Brassinosteroid Insensitive 1) and BR synthetic genes were reduced in ravl1 and were increased in overexpressors. However, RAVL1 did not exert any effect on the mRNA levels of BR catabolic genes (Je and Han, unpublished data). These data led to the conclusion that RAVL1 activates both BRI1 and BR synthetic genes. In vitro binding of RAVL1 to the E-box elements (CANNTG) of the BRI1, D2 (rice Ebisu Dwarf/Dwarf2), D11 (rice DWARF11) and BRD1 (BR-deficient Dwarf1/BR-C6 oxidase) promoters was demonstrated by EMSA (electrophoretic mobility shift assay) and a DNase footprinting assay. A transient transcription activation assay using protoplasts and ChIP (chromatin immunoprecipitation) using GFP fused RAVL1 overexpressors demonstrated that RAVL1 binds directly to the promoter regions of both BRI1 and BR synthetic genes (D2, D11 and BRD1) via E-box elements. Moreover, the differential affinity of RAVL1 for E-box elements was consistently observed in EMSA, DNase footprinting and transient expression assays.

Since RAVL1 mutants exhibit BR insensitivity and reduced expression levels of BRI1, it was very important to determine whether the loss of BR sensitivity or response in ravl1 plants was a consequence of the reduction in BRI1 mRNA. To address this question, two genetic strategies have been employed. One strategy was to analyze the phenotypes of osravl1 plants in the presence of a dominant activation allele of BRI1 (bri1-D),² and the other was to examine BRI1 mutant (bri1 d61-1)³ plants constitutively overexpressing RAVL1. Both experiments should equally address the question of whether the RAVL1 phenotypes are mediated via BRI1. In bri-D ravl1 double mutant plants, the BRI1 dominant activation allele restored the sensitivity of the ravl1 mutant to BR. In bri1 d61-1 and RAVL1 OX double mutants, the RAVL1 OX phenotype is greatly reduced by the presence of bri1 d61-1. In addition, the typical RAVL1 OX phenotypes were not detected in BR synthetic mutants (d2 and d11).^4,5 These data indicate that BR sensitivity is conferred by the action of RAVL1 on the BRI and BR synthetic genes. The next important inquiry was to define the relationship between RAVL1 and BZR1 (Brassinazole-Resistant 1), which is a transcription factor involved in BR feedback mechanisms.⁶ The activity of the BRD1 promoter,⁷ in which a BZR1-binding BRRE element is adjacent to a RAVL1-binding E-box element, was examined using transient transcription assays. The data showed that BZR1-mediated repression is dominant over RAVL1-mediated activation in regulating the target BR synthetic gene BRD1.

In summary, RAVL1 confers BR-sensitivity to plants by regulating the expression of BRI1 and the BR synthetic genes (Fig. 1). However, RAVL1-mediated activation can be counteracted by BR feedback repression of shared target genes. Significantly, the expression of RAVL1 is not affected by BR feedback regulation. Therefore, RAVL1 functions to maintain basal BR homeostasis, which ensures the appropriate basal levels of cellular BR. Since BR treatment did not affect the expression of RAVL1, auxin may be a candidate for the control of the RAVL1-mediated activation system. In fact, RAVL1 expression is repressed by NAA treatment. In the RAVL1::Ds gene trap line, the GUS reporter was downregulated by 2.5 µM NAA treatment (Je and Han, unpublished data). Recently, we isolated an ARF (Auxin Response Factor) that binds to the ARE (Auxin Response Element) of the RAVL1 promoter (Xuan and Han, unpublished data).

Figure 1.

Functional model of RAVL1 action in BR homeostasis. The conventional BR signaling circuit leading to feedback repression of synthetic genes by BZR1 is shown by the dotted lines. (1) RAVL1 mediates activation of both BRI1 and the BR biosynthetic genes, which have antagonistic actions on BR levels. (2) Activation of biosynthetic genes by RAVL1 may be repressed by BRZ1. Auxin may suppress RAVL1-mediated activation pathways.

The ravl1 mutants do not show severely defective phenotypes as do other BR transcription factors in Arabidopsis. They exhibit a mild semi-dwarf phenotype, which may be due to the functional redundancy of RAVL1. In fact, rice has a RAVL1 homolog, RAVL2, whose amino acid sequence shows 68% identity with RAVL1. Both proteins share sequence similarity not only in the B3 domain but also in the C-terminal region of RAVL1. However, the RAVL2 promoter does not contain ARE cis-elements. In Arabidopsis, a complex of the gene products of BES1 (Bri1 Ems Suppressor 1) and BIM1 (BES1-interacting Myc like 1), which carries a bHLH DNA binding domain, activates BR-related genes via E-box elements.⁸ Therefore, it will be important to determine whether BES1 and RAVL1 share common target genes. Additionally, identification of the genes activated by RAVL1 that can be repressed by the BR negative feedback loop will be important. The morphological and physiological roles of these genes should further advance our knowledge of BR-dependent mechanisms of plant development.

Addendum to: Je BI, Piao HL, Park SJ, Park SH, Kim CM, Xuan YH, et al. RAV-Like1 maintains Brassinosteroid homeostasis via the coordinated activation of BRI1 and biosynthetic genes in rice. Plant Cell. 2010;22:1777–1791. doi: 10.1105/tpc.109.069575.