Abscisic Acid Signal Transduction: Function of G Protein–Coupled Receptor 1 in Arabidopsis

GTP binding proteins, called G proteins, are important signaling molecules in eukaryotes. Heterotrimeric G proteins, which consist of three subunits, α, β, and γ, make up one of the major classes of G proteins involved in many signaling pathways in animals, including the perception of light, odor, certain tastes, hormones, and neurotransmitters (Cabrera-Vera et al., 2003). In animal systems, heterotrimeric G protein signaling is linked to signal perception by G protein–coupled receptors (GPCRs), a large class of membrane-bound receptors characterized by a conserved heptahelical, or 7-transmembrane, domain structure. GPCRs interact directly with the α subunit of heterotrimeric G proteins. In the absence of an activating ligand (agonist), the α subunit binds GDP and forms an inactive heterotrimeric complex with the β and γ subunits. Agonist binding to the receptor (signal perception) promotes the release of GDP. This in turn causes release of the β and γ subunits (which remain bound together as an undissociable heterodimer), thereby activating the free α subunit and βγ pair for further interactions with downstream effectors. Intrinsic GTPase activity of the α subunit catalyzes release of a phosphate from bound GTP, returning the G protein to its GDP-bound, heterotrimeric state.

Mammals contain ∼23 distinct Gα, six different Gβ, and at least 12 Gγ subunits, in addition to numerous GPCRs, making G protein signaling a highly versatile and prominent signaling mechanism (Jones and Assmann, 2004). By contrast, the Arabidopsis genome contains a single Gα gene, GPA1, a single Gβ gene, AGB1, two genes for putative Gγ subunits, AGG1 and AGG2, and, based on sequence similarity as the criterion, a single putative GPCR, GCR1. In plants, heterotrimeric G protein signaling has been linked to hormone responsiveness (mainly abscisic acid [ABA] and gibberellic acid, but also brassinolide and auxin) and cell division, principally through analyses of gpa1 and agb1 knockout mutants of Arabidopsis (Assmann, 2002; Jones, 2002; Coursol et al., 2003; Lapik and Kaufman, 2003; Ullah et al., 2003). GCR1 activity has also been linked to the cell cycle and ABA responsiveness (Colucci et al., 2002; Apone et al., 2003). So it is reasonable to hypothesize that GCR1 interacts with GPA1 in Arabidopsis and that ligand binding to GCR1 regulates heterotrimeric G protein signaling via GPA1 and the βγ dimer. However, these interactions have not been demonstrated previously.

In this issue of The Plant Cell, Pandey and Assmann (pages 1616–1632) provide strong evidence that Arabidopsis GCR1 interacts with GPA1 and regulates ABA signaling. Preliminary tests with the classic yeast two-hybrid assay failed to show an interaction between GCR1 and GPA1. However, the yeast two-hybrid assay is known to be inefficient and somewhat unreliable with membrane-bound proteins. Pandey and Assmann therefore made use of a modified split ubiquitin assay in yeast (Ludewig et al., 2003) that is designed for detection of in vivo interactions between membrane proteins. In this assay, one of two proteins to be tested is fused to the N terminus of ubiquitin and the other is fused to the ubiquitin C terminus, which is also fused with an artificial transcription factor, PLV (constructed from the viral factor VP16). The fusion proteins are expressed in yeast. The N and C termini of ubiquitin separately are not recognized and cleaved by ubiquitin-specific proteases. However, the assay is designed such that if an interaction occurs between the two test proteins, the N and C termini of ubiquitin will reconstitute functional ubiquitin. Proteolysis by ubiquitin-specific proteases will then cleave and release PLV from the C terminus, thereby activating specific reporter genes integrated into the yeast genome. Pandey and Assmann used this assay, with appropriate controls, to show that GCR1 interacts specifically with GPA1. The interaction was further confirmed with an in vitro antibody pull-down assay and an in planta assay in gcr1 mutant plants. For the in planta assay, gcr1 mutant plants were transformed with GCR1 fused to a FLAG epitope tag expressed under the control of a glucocorticoid-inducible promoter. After induction of transgene expression, anti-FLAG antibody was used to detect GCR1-FLAG in a protein complex precipitated from soluble and membrane protein fractions with anti-GPA1 antibody and vice versa.

Pandey and Assmann also investigated the function of GCR1 in Arabidopsis by identifying and analyzing gcr1 T-DNA insertional mutants. Two independent T-DNA insertional lines were identified as single insertion mutations of GCR1 based on the position of the T-DNA insertion in the 2nd or 3rd intron of the GCR1 open reading frame and by analysis of copy number. RT-PCR analysis of gene expression showed no detectable GCR1 transcript in total RNA extracts. The gcr1 mutant plants were found to be hypersensitive to ABA in many classic ABA responses. Compared with wild-type plants, the mutants showed increased ABA inhibition of root growth, higher levels of expression of a set of ABA- and stress-regulated genes, and improved drought tolerance (Figure 1). Experiments with isolated epidermal peels showed that gcr1 mutants were also hypersensitive to ABA in their stomatal responses. gcr1 mutants showed increased ABA inhibition of stomatal opening and increased ABA promotion of stomatal closing relative to the wild-type plants. Interestingly, gpa1 (Gα subunit) mutants show insensitivity to ABA inhibition of stomatal opening (i.e., the opposite phenomenon) (Wang et al., 2001), which suggests that GCR1 may function as a negative regulator of GPA1 and of ABA responsiveness in guard cells.

Figure 1.

gcr1 Mutant Plants Exhibit Hypersensitivity to ABA and Are More Drought Tolerant Than Wild-Type Arabidopsis.

The photographs show recovery of gcr1 mutant (top) and wild-type (bottom) plants from drought stress 2 d after rewatering after 12 d of withholding water.

The nature of the specific ligand that binds to GCR1 in the ABA response pathway remains unknown. Although it remains possible that GCR1 is an ABA receptor, it has not previously been identified as an ABA binding protein. If it were the sole ABA receptor, one would expect gcr1 knockout mutants to exhibit constitutive ABA responses. Instead, the knockout phenotype is characterized by hypersensitivity to ABA. In animal systems, most, if not all, GPCRs function as dimers or higher order oligomers (Breitwieser, 2004). GPCRs may form homodimers or heterodimers and may interact with members of quite distinct receptor families. This greatly increases the versatility of this signaling mechanism because of the large number of GPCRs encoded in animal genomes. For example, GPCRs represent ∼1% of human genes (Breitwieser, 2004). Although GCR1 is the only Arabidopsis gene with distinct similarity to canonical GPCRs, there are numerous genes encoding other transmembrane proteins that might conceivably interact with GCR1. Pandey and Assmann point to a possible analogy between GCR1 and the GPCR Smoothened found in animal systems. The interaction of Smoothened with heterotrimeric G proteins is regulated by an interaction with the 12-transmembrane domain protein Patched and further influenced by the Patched ligand Hedgehog. Patched does not show similarity to GPCRs but rather is similar to a channel or transporter (Frank-Kamenetsky et al., 2002). It is possible that GCR1 similarly interacts with a different membrane-bound ABA receptor and/or other transmembrane proteins.

Although many questions remain to be answered, not least of which are the nature of the ABA receptor or receptors and the mechanism by which GCR1 negatively regulates GPA1, the work of Pandey and Assmann illustrates the complex nature of ABA signal transduction and suggests that GCR1 is a component of an ABA perception and signaling complex.