Regulation of Brassinosteroid Signaling by a GSK3/SHAGGY-Like Kinase

See allHide authors and affiliations

Science  15 Feb 2002:
Vol. 295, Issue 5558, pp. 1299-1301
DOI: 10.1126/science.1065769


GSK3/SHAGGY is a highly conserved serine/threonine kinase implicated in many signaling pathways in eukaryotes. Although many GSK3/SHAGGY-like kinases have been identified in plants, little is known about their functions in plant growth and development. Here we show that the Arabidopsis BRASSINOSTEROID-INSENSITIVE 2(BIN2) gene encodes a GSK3/SHAGGY-like kinase. Gain-of-function mutations within its coding sequence or its overexpression inhibit brassinosteroid (BR) signaling, resulting in plants that resemble BR-deficient and BR-response mutants. In contrast, reduced BIN2 expression via cosuppression partially rescues a weak BR-signaling mutation. Thus, BIN2 acts as a negative regulator to control steroid signaling in plants.

BRs are a unique class of plant steroids that play important roles throughout the plant life cycle (1). Plants that are defective in BR biosynthesis or perception display characteristic mutant phenotypes that include a dwarf stature, delayed flowering and senescence, reduced apical dominance and male fertility, and photomorphogenesis in the dark (2–5). Previous genetic screens for BR-response mutants in Arabidopsis have led to the identification of only one BR-signaling gene, BRI1, which encodes a critical component of a membrane BR receptor (4–8). How BR signals are transduced inside plant cells is still a mystery.

To better understand the molecular mechanism(s) of the plant steroid signaling, we have analyzed various dwarf and semidwarf mutants that were collected during our previous genetic screen for bri1mutants (5) and have identified a new genetic locus,BIN2, whose gain-of-function mutations resulted in plants that display bri1-like phenotypes. Detailed genetic and physiological studies suggested that bin2 mutations might define a negative regulator in a BR signal transduction pathway important for plant growth (9).

We have cloned the BIN2 gene using a map-based cloning strategy. Fine mapping analysis delimited the BIN2 locus to a 13-kb fragment of a completely sequenced bacterial artificial chromosomal (BAC) clone that was mapped to the middle of chromosome IV between markers CH42 and AG (Fig. 1A). Sequence analysis of the three genes in the region from two independently isolated bin2 alleles identified mutations in a gene that encodes a previously reported cytoplasmic serine/threonine kinase, ASKη (10) (Fig. 1B). The catalytic domain of ASKη displays 70% sequence identity to those of the mammalian GSK3β (11) and the Drosophila SHAGGY protein kinase (12) (Fig. 1C). GSK3/SHAGGY kinases are a group of highly conserved serine/threonine kinases implicated in numerous signaling pathways, controlling metabolism, cell fate determination, and tissue patterning in organisms ranging from yeast to mammals (13). Homologs have been identified in several plant species and there are 10 GSK3/SHAGGY-like kinase genes in theArabidopsis genome (14). However, little is known about their functions in plant growth and development. A few molecular and biochemical studies have suggested possible roles for plant GSK3/SHAGGY-like kinases in osmotic stress response (15), wound signaling (16), and flower meristem patterning (17).

Figure 1

Molecular analysis of the BIN2gene. (A) Mapping of the BIN2 gene on chromosome IV. BIN2 was previously mapped between markers CH42 and AG (9). Three overlapping BAC clones, F28J12, F28A21, and F13C5, were known to cover this region (26). A published F21A28 sequence (AL035526) was used to generate new molecular markers N, 25, 47, 50, 60, 72, and S that were used to narrow theBIN2 gene to a 13-kb fragment that contains three predicted genes. Numbers of recombinants for each marker are shown on the top. (B) The structure of the BIN2 gene and the positions of two bin2 mutations. The BIN2 gene is composed of 10 exons (black box) and 9 introns (open box). The start ATG and the termination codons are indicated, and the positions of the two bin2 mutations are relative to the start ATG codon. (C) Sequence alignment of the BIN2 protein with the human GSK3β and the Drosophila SHAGGY kinases. The identical amino acids are boxed (19). The positions and molecular nature of the two bin2 mutations are indicated.

Because the two bin2 alleles that we have isolated contain gain-of-function mutations (9), we could not perform a rescue experiment to further prove that the mutated kinase gene is indeed the BIN2 gene. Instead, we cloned a 6.2-kb genomic fragment containing the complete ASKη gene, recreated the bin2-1 mutation by site-directed mutagenesis (18), and transformed the mutated gene into the wild-typeArabidopsis plants. As shown in Fig. 2A, such a mutatedASKη gene recapitulated the phenotype of the homozygous bin2 mutant, providing proof that theASKη gene is indeed the BIN2gene. In agreement with its gain-of-function nature, a second site mutation in which Lys 69 was replaced by Arg (K69R) (19) that was known to destroy the GSK/SHAGGY kinase activity (20) eliminated the phenotypic effect of thebin2-1 mutation, although the K69R mutation itself had no effect on the morphology of the transgenic plants (Fig. 2A).

Figure 2

bin2-1 is likely a hypermorphic mutation. (A) Recapitulation of the bin2phenotype by a mutated ASKη gene containing the bin2-1 mutation. Shown here (from left to right) are a wild-type Arabidopsis seedling, transgenic plants expressing the wild-type ASKη gene, or mutatedASKη gene containing bin2-1, K69R, or bin2-1, K69R double mutation, and bin2homozygous mutant. (B) The bin2-1 mutation leads to a more active BIN2 kinase in E. coli. Transphosphorylation activities toward a GSK3β peptide substrate of the wild-type GST-BIN2 fusion protein and its three mutant counterparts containing bin2-1, K69R, or bin2-1, K69R double mutation were measured, by using an in vitro phosphorylation assay (21), and expressed as a percentage of the specific activity of the wild-type GST-BIN2 fusion protein. Average activity (n = 3) and standard deviation are indicated. (C and D)BIN2 overexpression gives rise to an allelic series ofbin2 phenotype. (C) Seedlings grown on synthetic medium; (D) the same plants grown in soil under similar growth conditions. (E) Analysis of BIN2 expression of theBIN2 transgenic plants. Total RNAs (5 μg each) isolated from wild-type plants, bin2-1, and BIN2transgenic plants were separated by a formaldehyde-containing agarose gel, blotted to nylon membranes, and hybridized first with a full-length BIN2 probe and then with an 18S rDNA probe.

To investigate the biochemical effect of the bin2-1mutation, we expressed the wild-type BIN2 and its three mutant counterparts containing bin2-1, K69R, and bin2-1, K69R double mutations, respectively, as glutathioneS-transferase (GST) fusion proteins in Escherichia coli and measured their kinase activities using a synthetic peptide that contains only one unphosphorylated serine residue as a substrate (21). As indicated in Fig. 2B, although both K69R mutated proteins display almost no kinase activity, the GST-BIN2 fusion protein containing the bin2-1 mutation alone shows a statistically significant increase (33 ± 12%,n = 3) in kinase activity over the wild-type fusion protein, suggesting that bin2-1 might be a hypermorphic mutation. If this were true, we would expect that overexpression of the wild-type BIN2 gene driven by its native promoter, a “transgenic hypermorph,” could also lead to the bin2phenotype. Indeed, ∼10% of the transgenic plants containing the wild-type BIN2 genomic construct exhibited an allelic series of bin2 phenotypes (Fig. 2, C and D). Northern analysis with total RNAs isolated from representative transgenic plants revealed a correlation between BIN2 expression level and phenotypic severity (Fig. 2E). BIN2 overexpression can lead to increased biosynthesis of BIN2 proteins, resulting in an increase in BIN2 kinase activity. In both cases, an increased BIN2 kinase activity could phosphorylate and inactivate more positive BR signaling proteins to block BR signal transduction, giving rise to a bri1-like phenotype.

To further prove that BIN2 acts negatively to regulate BR signaling, we generated transgenic plants in a weak bri1 mutant background [bri1-301 (22)] using either a sense or antisense BIN2 cDNA driven by the strong, constitutively active cauliflower mosaic virus 35S promoter. If BIN2 acted as a negative regulator in BRI1-mediated BR signaling, we would expect that a reduction in BIN2 expression could, at least partially, suppress the weak bri1 mutation, whereas its overexpression could lead to a stronger bri1 phenotype. Although we did not observe any phenotypic alteration with the35S-BIN2 antisense construct, we identified three phenotypic classes of transgenic plants containing the 35S-BIN2 sense construct, including plants resembling the parental bri1-301mutant, lines showing a suppressed bri1 phenotype, and extreme dwarfs similar to bri1-101, a strong bri1mutant (Fig. 3, A to C). We suspected that the enhanced bri1 phenotype was likely caused by BIN2 overexpression driven by the strong 35Spromoter, whereas the suppressed bri1 phenotype was due to a reduced BIN2 expression caused by cosuppression.

Figure 3

A reduction in BIN2 gene expression partially suppressed a weak bri1 mutation. (A) Seedlings grown on synthetic medium [from left to right in (A to D)], wild-type, bri1-101, bri1-301, suppressed, and enhanced bri1-301 transgenic plants containing the 35S-BIN2 cDNA transgene). (B) Four-week-old seedlings grown in soil under similar growth conditions. (C) Leaf shape comparison. (D) Northern analysis of BIN2 gene expression. Total RNAs (5 μg each) isolated from wild-type plants, bri1-101, bri1-301, suppressed bri1 plants, and extreme dwarf mutants were separated by a formaldehyde-containing agarose gel, blotted to a nylon membrane, and hybridized first with a full-length BIN2 probe (probe 1 to detect the total BIN2 messages), and then with probe 2 (derived from the 3′-untranslated region of the BIN2 cDNA to detect the endogenous BIN2 messages). After stripping off probe 2, the same filter was hybridized with an 18S rDNA probe to control equal loading of total RNAs for each sample. Hybridization signals were quantified by a Bio-Rad GS-363 Molecular Imager System and normalized relative to respective 18S rRNA signals. The numbers listed in (D) are relative signal strength calculated as a ratio between a normalizedBIN2 signal of a given sample and the normalizedBIN2 signal of the wild-type plants.

To test this prediction, we performed a Northern blot analysis with total RNAs isolated from wild-type plants, the two bri1mutants, one representative each of the suppressed and enhancedbri1-301 transgenic plants containing the35S-BIN2 transgene. Using a full-length BIN2 cDNA as a probe, we observed a ∼50-fold accumulation of BIN2transcripts in the enhanced bri1-301 line compared with the wild-type control (Fig. 3D). However, no difference in hybridization signal was observed between the suppressed line and the other three tested plants, most likely due to cross-hybridization of the full-length BIN2 probe with other Arabidopsis GSK3/SHAGGY-like genes. To see if the suppressedbri1-301 line accumulated fewer BIN2 messages, we used a gene-specific probe corresponding to the 3′-untranslated region of the BIN2 cDNA that was not included in the35S-BIN2 transgene. As shown in Fig. 3D, the suppressed line accumulated ∼20% of the amount of BIN2 transcripts in the wild-type control. Because such a cosuppressed line is equivalent to a reduction-of-function bin2 mutant, this result provides a further support for BIN2 being a negative regulator in a BRI1-mediated BR signaling pathway.

Our results indicate that a GSK/SHAGGY-like kinase plays an important role in a steroid signal transduction pathway. In animals, GSK3/SHAGGY is a constitutively active kinase that negatively regulates a variety of substrates by phosphorylation, and the inhibition of its kinase activity constitutes a key event in many signal transduction pathways (13). It is well established that GSK3/SHAGGY kinase itself is negatively regulated by protein phosphorylation or protein-protein interaction in response to a variety of stimuli (13). If BIN2 functions as a negative regulator in BR signaling, a crucial step in this signal transduction pathway would be to inhibit the BIN2 kinase activity in response to BR signals, which are thought to be perceived by BRI1, a leucine-rich-repeat–receptor–like kinase (5–8). Thus, it is quite tempting to speculate that BRI1 might be one of the upstream kinases that could phosphorylate and inactivate BIN2 to relieve its inhibitory effect on downstream BR signaling components. However, repeated experiments using yeast two-hybrid analysis and in vitro protein interaction assay failed to demonstrate a direct physical interaction between BRI1 and BIN2, suggesting the existence of additional regulators that are involved in suppressing BIN2 activity. We hypothesize that, in the absence of BRs, BIN2 is a constitutively active kinase that can phosphorylate and inactivate positive BR signaling proteins to block BR signal transduction. The identification of upstream regulators and downstream targets of the BIN2 kinase is crucial to understand the molecular mechanisms of steroid signaling in plants.

  • * To whom correspondence should be addressed. E-mail: jian{at}


View Abstract

Navigate This Article