Special Perspectives

Brassinosteroid Signaling: A Paradigm for Steroid Hormone Signaling from the Cell Surface

See allHide authors and affiliations

Science  01 Dec 2006:
Vol. 314, Issue 5804, pp. 1410-1411
DOI: 10.1126/science.1134040

Abstract

Plants use the coordinated action of several small-molecule hormones to grow and develop optimally in response to a changing environment. Among these hormones are the brassinosteroids (BRs), the polyhydroxylated steroid hormones of plants. BRs bind a small family of leucine-rich repeat receptor kinases at the cell surface, thereby initiating an intracellular signal transduction cascade that results in the altered expression of hundreds of genes.

Brassinosteroids (BRs) are small growth-promoting molecules found at low concentrations throughout the plant kingdom. In a technical tour de force, researchers in the late 1970s purified a bioactive steroid from Brassica napus bee-collected pollen. The most active BR was identified by single-crystal x-ray analysis as a steroidal lactone and was named brassinolide (BL) (1). The role of BRs as plant hormones was clarified in the mid-1990s with the discovery of Arabidopsis thaliana mutants that were deficient in BR biosynthesis. BR-deficient mutants are extremely dwarfed with very small curled leaves; the mutants can be rescued to wild-type stature by exogenous application of BL (2, 3) (Fig. 1). Analysis of these mutants showed that BRs play a role in cell expansion and division, differentiation, and reproductive development. The regulation of these processes by BRs allows plants to develop a body plan that is optimal for their ambient environment and may confer increased adaptation to various stresses. The ease of analysis of the BR-deficient phenotype allowed the identification of the plasma membrane–localized BR receptor and much of the intracellular signaling pathway that regulates BR-responsive gene expression (46). However, key open questions remain with respect to how the BR signaling pathway contributes to the adaptive plasticity of plant growth.

Fig. 1.

A model of BR control of Arabidopsis size. Arabidopsis biosynthetic mutants that do not produce brassinolide (BL) are dwarf (upper left photo) but can be rescued to full stature by exogenous application of BL (upper right photo). BR signaling mutants (lower left photo) cannot be rescued by exogenous BL. In the absence of BRs, the kinase domains of the BRI1 homodimer are inhibited by both their own C-terminal domain and by an interaction with BKI1. This allows the GSK3 homolog, BIN2, to phosphorylate and inactivate the brassinosteroid response transcription factors (BRFs), including BES1 and BZR1. Direct binding of BL to BRI1 homodimers results in conformational changes of the kinase domain, leading to the phosphorylation of the C-terminal domain of BRI1 and phosphorylation of BKI1, which causes displacement of BKI1 from the plasma membrane and the release of autoinhibition of BRI1. These events lead to BRI1's association with BAK1 and consequent activation of the receptor. The active signaling receptor complex inhibits the activity of BIN2 by an unknown mechanism, allowing dephosphorylation of the BRFs by BSU1 and activation or repression of their target genes and optimal plant growth (lower right photo). The horseshoe-shaped representations of BRI1 and BAK1 LRR domains, as well as the putative LRR (red domains) interactions, are inferred from structural models of LRR-containing proteins. The atypical LRR21 is represented by a yellow domain. The BL docking into the binding site is speculative. Phosphorylation events are indicated by a circled P.

BRs have structures similar to those of animal steroid hormones. BR biosynthesis enzymes share sequence identity with mammalian steroid biosynthetic enzymes (e.g., DET2 is a plant ortholog of mammalian steroid 5α-reductases, and most of the other biosynthetic genes are cytochrome P450s) (7). The major branch of the biosynthetic pathway, from campesterol to BL, was determined using a combination of Arabidopsis genetics and feeding experiments with BR biosynthetic intermediates.

Forward genetic screens for dwarf mutants that were not rescued by exogenous addition of BL led to the identification of multiple mutant alleles of a single locus, BRI1 (brassinosteroid insensitive 1) (8, 9) (Fig. 1). BRI1 is a leucine-rich repeat receptor kinase (LRR-RK) that has an extracellular domain with an N-terminal signal peptide followed by 24 imperfect leucine-rich repeats (LRRs), a single transmembrane domain, and an intracellular serine-threonine kinase domain followed by a short C-terminal tail. BRI1 is the major BR-binding activity of Arabidopsis. The minimal BR-binding domain is a 94–amino acid subdomain that includes the 70–amino acid “island” just proximal to LRR20 and the atypical LRR, LRR21 (Fig. 1) (10). As such, the signaling pathway defined by BRI1 represents a paradigm for steroid perception at the cell surface and may have implications for the understanding of the growing field of plasma membrane steroid signaling in metazoans.

In the absence of steroid, the kinase activity of BRI1 is inhibited by both cis and trans mechanisms. BKI1 (BRI1 kinase inhibitor 1) is a plasma membrane–associated phosphoprotein that interacts directly with the kinase domain of BRI1 (6, 11). Binding of BRs to preformed BRI1 homooligomers triggers the rapid dissociation of BKI1 from the plasma membrane (11). In vitro, BKI1 interferes with the interaction of BRI1 with its signaling partner, a second plasma membrane–localized LRR-RK called BAK1 [BRI1-associated receptor kinase 1, also known as SERK3 (somatic embryogenesis receptor kinase 3)] (12).

Although the precise sequence of events is not clear, BR binding also allows autophosphorylation of critical serine and threonine residues located in the activation loop of the kinase domain of BRI1 (13, 14). This, in turn, alleviates the autoinhibitory effect of the C-terminal tail on kinase activity and allows further autophosphorylation of the receptor, which increases the affinity of BRI1 for BAK1 (13). BAK1 has a short extracellular domain comprising five LRRs. Although BR binding to BRI1 is independent of BAK1, a BRI1-BAK1 heterooligomer may be the active signaling complex. SERK1, a BAK1 homolog, also interacts with BRI1, and genetic experiments have implicated this protein in the signaling pathway (12). In addition to the plasma membrane, a signaling-competent BRI1-BAK1 hetero-oligomer was detected in bona fide endocytic compartments of plant protoplasts (12). Furthermore, overexpression of BAK1 appears to augment the internalization of BRI1 in heterologous cell cultures. Thus, the function of the SERK family in BR signaling may be to facilitate the entry of BRI1 into these intracellular compartments. The physiological role of an endosome-localized BRI1-BAK1 hetero-oligomer is not currently known.

Only the BRI1 locus was defined by loss-of-function mutations that caused BR-resistant dwarfism; however, several additional components of the pathway were identified by analysis of gain-of-function phenotypes or plants with increased sensitivity to BRs. A bin2 (brassinosteroid insensitive 2) dwarf mutant phenotype results from a semi-dominant mutation. BIN2 is one member of an Arabidopsis subfamily of glycogen synthase kinase 3 (GSK3, also known as Shaggy-like kinases). Although single loss-of-function alleles revealed no effect on BR signaling (6, 15), reduced expression of the entire subfamily results in plants with enhanced BR responses, supporting a role for these three kinases as negative regulators of BR signaling (15). BIN2 localizes to multiple subcellular compartments but appears to exert its largest effects on BR signaling when it is retained in the nucleus.

BSU1 (BRI1 suppressor 1) was identified as a dominant suppressor of a weak bri1 mutant (4, 6). BSU1 overexpression substantially suppresses the dwarf phenotypes associated with either the bri1 or bin2 mutations, which suggests that BSU1 is a positive regulator of BR signaling that acts on the same process or downstream of BIN2. BSU1 encodes a nuclear-localized serine-threonine phosphatase with an N-terminal domain comprising Kelch repeats. Loss of BSU1 function has no effect on Arabidopsis, but when the expression of BSU1 and three related genes is reduced by RNA interference, the resulting plant is dwarfed (16, 17). The substrates of BIN2 and BSU1 are likely to be a family of plant-specific transcription factors. The founding members of this family, BES1 (bri1 ems 1) and BZR1 (brassinazole resistant 1), are 89% identical and contain multiple predicted GSK3 phosphorylation sites (4, 15). Recent models propose that the balanced activities of BIN2 and BSU1 directly control the phosphorylation state of BES1 and BZR1. BIN2-induced phosphorylation of BES1 inhibits its transcriptional activity through impaired multimerization and DNA binding activity at BR-responsive target promoters. In contrast, in the presence of BRs, BES1 and BZR1 are dephosphorylated by the combined inactivation of BIN2 and the phosphatase activity of BSU1, which allows them to homo- or heterodimerize and bind more efficiently to the BR-responsive elements to either positively or negatively regulate BR-responsive target genes (15, 18).

Despite tremendous progress in understanding the molecular and cellular effects of BRs, key issues remain unanswered. First, details of the receptor activation mechanism need to be clarified and the physiological role of the endosomal BRI1-BAK1 hetero-oligomer needs to be determined. Second, a major gap exists in the pathway between events at the plasma membrane and in the nucleus. Understanding how BIN2 is inactivated and how its subcellular localization may be altered in response to BRs may help fill in this gap. A third area of investigation revolves around the pleiotropic actions of BRs on plant tolerance to temperature, salt, and pathogens. Although the signaling pathway defined by BRI1 is known mostly for controlling the rapid increase in tissue mass, it remains possible that the other effects of BRs are mediated by different signaling pathways. Finally, every signaling component, from BRI1 to the regulated transcription factors, is redundantly encoded in the genome. Perhaps redundancy in BR signaling could increase the robustness of the pathway to mutations; alternatively, partial redundancy due to overlapping expression patterns of signaling components may help to fine-tune the pleiotropic BR response. A systematic analysis of the function and expression patterns of family members will help to unravel these questions.

Ultimately, the goal of this work is to understand how plant size is controlled and to be able to manipulate plant growth for human benefit. BRs play an important role in the processes that control plant size, but other small-molecule hormones also contribute to cell expansion and division. Although there is redundancy within a given hormonal pathway, loss of response to any one plant hormone cannot be compensated for by the action of another hormone. To add another level of complexity, the entire program can be altered by changes in the environment's ambient light or temperature. Plants provide an accessible model for answering questions of organ or organismal size because the number of cell types and different organs is small, organ size is easily manipulated by environment, and many of the individual signaling pathways are known. Questions that can be addressed include how the levels of plant hormones change throughout development and in response to an ever-changing environment, and how these different pathways interact within individual cells. Ultimately, we may answer the age-old question of how the size of an organism is determined.

References and Notes

Stay Connected to Science

Navigate This Article