Plant G Proteins, Phytohormones, and Plasticity: Three Questions and a Speculation

See allHide authors and affiliations

Science's STKE  21 Dec 2004:
Vol. 2004, Issue 264, pp. re20
DOI: 10.1126/stke.2642004re20


Heterotrimeric guanine nucleotide-binding proteins (G proteins) composed of Gα, Gβ, and Gγ subunits are important transducers of hormonal signals in organisms as evolutionarily distant as plants and humans. The genomes of diploid angiosperms, such as that of the model species Arabidopsis thaliana, encode only single canonical Gα and Gβ subunits, only two identified Gγ subunits, and just one regulator of G protein signaling (RGS) protein. However, a wide range of processes—including seed germination, shoot and root growth, and stomatal regulation—are altered in Arabidopsis and rice plants with mutations in G protein components. Such mutants exhibit altered responsiveness to a number of plant hormones, including gibberellins, brassinosteroids, abscisic acid, and auxin. This review describes possible mechanisms by which such pleiotropic effects are generated and considers possible explanations for why G protein component mutations in plants fail to be lethal. A possible role of G protein signaling in the control of phenotypic plasticity, a hallmark of plant growth, is also discussed.


G proteins are found in organisms as diverse as slime molds, sponges, and humans. In mammals, G proteins mediate signals from various sensory stimuli and neuroendocrine ligands; to do so, they draw upon a large stock of distinct G protein–coupled receptors (GPCRs) and regulatory proteins (1, 2). It is estimated that about half of today’s pharmaceuticals target G protein–based pathways (3), and G protein component defects have been linked to various genetic diseases (4, 5).

New studies are also revealing roles of G proteins in hormonal signaling in flowering plants. However, the textbook definition of a hormone as a molecule that is produced by one specific organ and conveyed to target tissues, where it elicits a physiological response at low concentration, does not hold true for plant hormones. Most plant hormones are synthesized by several different tissues or cell types, and can act locally as well as at a distance. Like hormones of mammals (2), plant hormones can be perceived by multiple types of cells, and the nature of the responses that are thereby engendered can differ greatly between one cell type and another.

In this review, the roles of G proteins in plant hormonal signaling and response are discussed in the context of three questions:

1) What are the plant hormonal responses that are influenced by G proteins?

2) How can only two plant heterotrimers mediate so many plant hormone responses?

3) Given the paucity of plant G protein subunits, why aren’t plant G protein component knockout mutations lethal?

Some speculation is also offered regarding the importance of G proteins in the context of phenotypic plasticity and evolutionary fitness. Discussion is limited to the two model plant species for which the full sequence of the genome is available, Arabidopsis thaliana and rice (Oryza sativa), and to the specific realm of hormonal responses. Plants respond to a remarkable diversity of environmental cues and metabolites (Fig. 1), and G protein involvement in other aspects of plant signaling—notably responses to pathogens, pollutants, and drought—has also been demonstrated. For other aspects of plant G protein function, including discussions of phylogeny and protein structure, the reader is referred to several recent reviews (68). In particular, Jones and Assmann (9) provide a comparative evaluation of plant versus mammalian G protein signaling components and downstream effectors.

Fig. 1.

Stages of development of the model plant Arabidopsis thaliana. Hormones and metabolites that affect physiology, growth, and development are listed. Figure influenced by figure 1 of (77) and figure 1 of (80).

Background and Overview: The Plant Heterotrimer and Associated Regulatory Proteins

In the standard paradigm of G protein signaling, Gα-GDP, Gβ, and Gγ form an inactive complex. Ligand binding to an associated GPCR elicits a conformational change in the Gα subunit, leading to exchange of GDP (guanosine diphosphate) for GTP (guanosine triphosphate) and a consequent conformational change that releases the Gβγ dimer. Free Gα and Gβγ are available for interaction with downstream effectors until intrinsic GTPase activity of the Gα subunit allows reunion of the heterotrimer. Experimentally, G protein activation also can be induced by mutagenesis of particular amino acids in the Gα subunit, resulting in the creation of a permanently active Gα conformation, or by application of nonhydrolyzable GTP [for example, guanosine 5′-O-(3′-thiotriphosphate) (GTP-γ-S)], which locks the G protein in the active state (1).

Recent pharmacological studies have revealed that other states of the GPCR beyond ligand-activated and inactive are possible. Many GPCRs have basal activity that can be altered by agonists and antagonists, as well as by inverse and protean agonists. Protean agonists either stimulate or inhibit receptors, depending on the initial state of the receptor (1012) (Fig. 2). Posttranslational modifications of GPCRs also modulate signaling, as does GPCR homo- or heterodimerization (1, 10). Thus, it may be appropriate to model GPCRs in the same way that electrophysiologists model ion channels, which can be open (active), closed (not active), or inactive (in a state not susceptible to activation) (Fig. 2). Adding further layers of complexity, Gα activation by proteins that do not bear the archetypal seven transmembrane–spanning (7TMS) structure of the GPCRs occurs in some signaling pathways (13, 14), and signaling by some GPCRs to non-Gα proteins also occurs (15).

Fig. 2.

Known modes of GPCR modulation in mammals, by analogy to ion channel models. Ion channel models commonly consider states of the ion channel as open (O, active), closed (C, not active), and inactive (I, a state in which the channel cannot be activated, even in the presence of an activating stimulus). Similar reasoning can be applied to GPCRs. In the absence of antagonists, agonists increase activity beyond the basal level (O1); inverse agonists decrease activity beneath the basal level. Protean agonists can have either agonist or inverse agonist effects, depending on the level of constitutive activity. Antagonists bring activity back to the basal level. Localization, modification, and interaction with other proteins also affect GPCR status.

Because the plant G protein field is relatively new, the complex regulation of GPCRs described above has yet to be experimentally evaluated in plants. However, basic elements of G protein signaling appear to be conserved between plants and animals, although plant genomes encode far fewer heterotrimeric subunits than the 23 Gα (including splice variants), 5 Gβ, and about 11 Gγ subunits present in mammalian genomes (16). Indeed, in the fully sequenced Arabidopsis and rice genomes, the Gα and Gβ subunits are both encoded by single-copy genes, and there appear to be only two Gγ subunits (Table 1). Respectively, these genes are named GPA1 in Arabidopsis (RGA1 in rice), AGB1 (RGB1 in rice), and AGG1 and AGG2 (RGG1 and RGG2 in rice) (1723). All four of the Arabidopsis subunits exhibit ubiquitous expression [reviewed in (7)]. Existence in the plant plasma membrane of a genuine heterotrimer has been documented by gel filtration analysis of protein complexes in rice; moreover, dissociation of this complex by nonhydrolyzable GTP has been observed (23). Structural modeling of the Arabidopsis subunits suggests that an analogous heterotrimer exists in Arabidopsis (24). Plant Gαs exhibit intrinsic GTPase activity, albeit at a lower rate than that of mammalian Gαs (2528), and constitutively active mutants can be generated by the same strategy as in mammalian systems (24, 27). Downstream effectors of G protein signaling in plants include phospholipase C, K+ channels, anion channels, and Ca2+ channels (2931), although the number of elements situated between the G protein and each of these effectors remains unknown (9). GPA1 has been shown biochemically to interact with and thereby inhibit the activity of phospholipase D1α (PLD1α); GTP relieves this inhibition (32). A cupin-domain protein, AtPirin1, interacts with GPA1 in a yeast two-hybrid assay (33). It is expected that many more effectors will be uncovered.

Fig. 3.

Models consistent with G protein component mutant phenotypes in hormonal regulation of seed germination. (A) A GPCR directly couples perception of GA to activation of the G protein heterotrimer in a high-sensitivity GA-sensing pathway, whereas a low-sensitivity GA-sensing pathway uses a different, non–GPCR-based pathway. (B) The heterotrimer does not couple perception of GA but strengthens the promotive effects of BR on GA-stimulated germination. (C) The heterotrimer does not directly couple to any hormone perception pathway, but its status modulates the strength of both stimulatory (for example, GA and BR) and inhibitory (for example, glucose and ABA) inputs into seed germination. In all panels, dotted lines represent multistep pathways.

Fig. 4.

Hypothetical examples of how phenotypes of wild-type plants (blue boxes) and G protein component mutant plants (green boxes) might differ in terms of either plasticity (represented by shifts in the vertical position of the box across environments) or phenotypic variation (represented by the dimensions of the box within an environment). The "ideal" phenotype is arbitrarily assumed to be the center point of the wild-type phenotypic response in each environment. (A) Wild-type and mutant plants exhibit the same amount of phenotypic variation within an environment, but the mutant has no phenotypic plasticity. (B) Wild-type and mutant plants exhibit the same amount of phenotypic plasticity, but the mutant exhibits reduced phenotypic variation within each environment. As illustrated, the mutant is converging on the ideal phenotype and may thus be predicted to exhibit greater fitness. (C) Wild-type and mutant plants exhibit the same amount of phenotypic plasticity, but the mutant exhibits increased phenotypic variation within each environment as well as divergence from the "ideal" phenotype. (D) Wild-type and mutant plants exhibit the same amount of phenotypic plasticity as well as the same amount of variation, but the mutant’s phenotypic response curve (reaction norm) is shifted away from the "ideal" phenotype.

Table 1.

Plant G protein components discussed in this review.

Plants also appear to have far fewer GPCRs than the 800+ complement found in mammals (34). In fact, only one Arabidopsis protein, GCR1, bears prominent sequence similarity (within a limited region) to metazoan GPCRs (3537). GCR1 can be co-immunoprecipitated with GPA1 from plant tissue (36). This result is certainly suggestive of GPCR function for GCR1, but until a GCR1 ligand is found, its classification as a GPCR remains putative. Plants may have additional 7TM proteins that function as GPCRs but are difficult to discover via searches for sequence similarity, just as mammalian GPCRs within the different GPCR families (38) can have little or no sequence similarity to one another.

Proteins that modulate signaling through GPCRs and G proteins in mammals include (i) GPCR kinases (GRKs) and non–GRK-type receptor kinases, which desensitize GPCRs to activation by ligand; (ii) GPCR phosphatases, including protein phosphatase 2A enzymes (PP2As), which resensitize GPCRs; (iii) arrestins, which terminate signaling by promoting GPCR internalization (39) and also function as adaptor proteins that regulate mitogen-activated protein kinase cascades (40); (iv) phosducins, which interfere with signaling in certain tissues by scavenging Gβγ dimers (41); and (v) regulators of G protein signaling (RGS), which promote Gα GTPase activity and thus typically accelerate return to the heterotrimeric state (42). Of these regulatory proteins, PP2As (43) and one RGS protein (RGS1) (28) are present in Arabidopsis, as confirmed by both sequence and functional analysis. Arabidopsis RGS1 is unusual in that it has both an RGS domain and a 7TM domain; thus, it may function as both a GPCR and an RGS (28).

What Are the Plant Hormonal Responses Influenced by G Proteins?

Application of each of the major plant hormones leads to characteristic responses of various plant organs at particular stages of development. Although responses to exogenous hormones do not always faithfully reflect responses to endogenous hormone levels (44), these stereotypical responses have nonetheless become useful hallmarks for evaluation of hormone sensitivity or responsiveness, or both, in genetically altered experimental plants. Analysis of Gα, Gβ, GCR1, and RGS1 mutants for hormonal responses has been initiated in Arabidopsis, whereas in rice only mutants of the rice Gα subunit gene RGA1 have been studied to date. For both species, only a subset of plant hormones—primarily gibberellic acid, brassinosteroids, abscisic acid, and auxin—and a subset of the attendant responses have been characterized thus far.


Classical roles of gibberellic acid (GA) include promotion of seed germination, elongation of the hypocotyl [the portion of the seedling stem located below the cotyledons (Fig. 1)] and stem, leaf expansion, flowering, and fruit expansion. Indeed, growers spray some varieties of table grapes with GA to increase grape berry size (45). Growth-promotive effects of GA in the vegetative plant are typically attributed to induction of cell expansion rather than stimulation of cell division.

Arabidopsis mutants in the G protein α or β subunit (gpa1, agb1), the putative GPCR (gcr1), and the RGS protein (rgs1) all show reduced seed germination in response to exogenous GA application as well as increased sensitivity to inhibition of germination by the GA synthesis inhibitor paclobutrazol, whereas cell lines overexpressing the Gα subunit GPA1 are hypersensitive to GA in seed germination assays (37, 46). These results suggest that the heterotrimer either transduces or modulates the GA response of seeds; these possibilities are discussed in more detail below.

Arabidopsis seedlings expressing a constitutively active version of GPA1 (GPA1QL) have longer hypocotyls when grown in darkness as a result of increased cell elongation (28), which is also consistent with a GA-hypersensitive phenotype, although this has yet to be specifically tested. rgs1 mutants also show this long hypocotyl phenotype, consistent with the idea that RGS1 reduces the GPA1 activation state in wild-type plants (28).

During the final, reproductive phase of growth, Arabidopsis plants harboring G protein component mutations also show phenotypes suggestive of, but not yet definitely linked to, altered GA signaling. Thus, agb1 plants exhibit shorter floral parts and thicker, blunted fruits, a response that is somewhat phenocopied by overexpression of GPA1 (24, 47). Such similarity may occur because overexpression of GPA1 results in sequestration of AGB1 by GPA1, resulting in an AGB1-deficient phenotype.

In rice, the d1 mutant, harboring a null mutation of the RGA1 Gα subunit gene (23), also shows reduced GA sensitivity (4850). d1 mutants require higher concentrations of GA to induce expression of several GA-specific genes in the seed (50). In d1 seeds the aleurone layer, a tissue rich in hydrolytic enzymes, exhibits lower activity of α-amylase, a GA-induced enzyme that is crucial for mobilization of the carbohydrate reserves that nurture the embryo during germination (50). The rice mutants of RGA1 also show GA-related alterations in vegetative morphology; these mutants are dwarves as a result of decreased sensitivity to GA promotion of internode elongation (49, 50). However, other organs of rice, specifically the lower part or sheath of the second leaf, show normal sensitivity to GA (50), indicating tissue specificity of response.


Brassinosteroids (BRs) have only relatively recently received recognition as genuine plant steroid hormones (51). Exogenous BRs can stimulate cell expansion and have general growth-promotive effects on above-ground plant parts, although root growth can be inhibited (51). Brassinosteroid mutants exhibit a dark green phenotype when light-grown and exhibit aspects of light-grown morphology when grown in the dark. Application of the synthetic BR brassinolide (BL) promotes Arabidopsis seed germination in plants with mutations in genes related to GA biosynthesis or perception.

Only a little research has been done concerning BR responses in G protein component mutants. The rounded-leaf or "rotundifolia" phenotype of gpa1 and agb1 Gα and Gβ subunit mutants is similar to that observed in a cytochrome P450 mutant posited to be defective in plant steroid synthesis, suggesting a possible link (52). In Arabidopsis seeds pretreated with the GA biosynthesis inhibitor paclobutrazol, BL rescue of seed germination is reduced in gpa1, agb1, and gcr1 single mutants relative to the response of wild-type seeds (37, 46). Likewise, in the rice d1 mutant, BL stimulation of gene expression is less effective than in wild-type plants (53). These results suggest that disruption of G protein signaling reduces sensitivity to BR.

Abscisic Acid

In contrast to GA and BR, abscisic acid (ABA) has a general growth-retarding role. ABA is important for embryo development within the seed and for subsequent development of seed dormancy; indeed, some ABA-deficient mutants of maize show viviparous seed germination on the ear (54). In young seedlings under conditions where sufficient water is available, application of ABA delays overall growth and inhibits elongation of the primary (main) root. ABA also has important acclimatory roles in plant stress responses, particularly responses to drought, cold, and salinity (55). Throughout the plant body, elevated concentrations of ABA increase transcript abundance of a number of genes whose products are important for stress tolerance. ABA also contributes to plant water retention during drought by inhibiting the opening and promoting the closure of stomata, microscopic pores on the leaf surface through which plants simultaneously take up CO2 and lose water vapor (Fig. 1) (56).

ABA responses have been evaluated in Arabidopsis G protein component mutants but not in corresponding mutants in rice. Arabidopsis gcr1 mutants exhibit ABA hypersensitivity in inhibition of primary root elongation and in expression of some ABA-related genes in whole seedlings (36). Arabidopsis gcr1 mutants are also hypersensitive to ABA inhibition of germination (36), a phenotype that is shared by single mutants of gpa1 and agb1 (33, 46) as well as by double and triple mutant combinations among these three genes (57).

However, with respect to stomatal aperture regulation, a quite different story is emerging. gpa1 and agb1 mutants show wild-type ABA induction of stomatal closure but actually exhibit insensitivity to ABA inhibition of stomatal opening, rather than the hypersensitivity displayed by other tissues (30, 58, 59). The similar phenotypes of the Gα and Gβ mutants suggest that either Gα or the concerted action of Gα and Gβγ mediate the ABA response in wild-type cells. Stomatal opening is driven in part by an influx of K+ ions into the two guard cells that border and define the stomatal pore. As a result of this influx, the cells take up water, swell, and separate from one another, thus widening the stomate. K+ influx occurs through inwardly rectifying K+ channels that are inhibited by ABA in wild-type plants but not in gpa1 [or agb1 (59)] mutants (30, 58). K+ influx occurs upon membrane hyperpolarization, and this driving force for K+ uptake is lessened by membrane depolarization. In wild-type guard cells, two processes that contribute to ABA promotion of depolarization are activation of anion channels that mediate anion efflux and activation of Ca2+-permeable channels that mediate Ca2+influx. Regulation of slow anion channels by ABA is altered in gpa1 mutants (30) and ABA-activation of Ca2+-permeable channels is abrogated (31), thereby contributing to ABA insensitivity of stomatal opening.

In guard cells, ABA activates sphingosine kinase activity, which leads to production of the lipid metabolite sphingosine-1-phosphate (S1P). In mammalian systems, S1P signals through specific GPCRs (60). S1P mimics ABA effects on K+ and anion channel regulation in wild-type plants but fails to do so in gpa1 mutants, which suggests that an S1P signal (generated by ABA) signals via G protein–based cascades in plants as well as in mammals (58).

Given the reduced ABA sensitivity of guard cells in gpa1 and agb1 mutants, one might then predict that mutation of the putative GPCR, GCR1, would also lead to ABA hyposensitivity, but this is not the case. gcr1 mutant guard cells are similar to gcr1 seeds and primary roots in showing hypersensitivity both to ABA inhibition of stomatal opening and to ABA promotion of stomatal closure. gcr1 guard cells also exhibit hypersensitivity to S1P in these two processes. These results imply that GCR1 is a negative regulator of ABA signaling, because elimination of gcr1 in mutant plants results in ABA and S1P hypersensitivity (36).

Although the electrophysiology of ABA-regulated ion channels has yet to be investigated in cell types other than guard cells, the fact that gpa1 (and agb1) mutants are hypersensitive to ABA in seed and root tissue, but hyposensitive to the hormone in stomatal opening, clearly points to cell specificity in the G protein dependence of ABA signaling pathways. Such cell specificity is also illustrated in auxin responses.


Auxin regulates both cell division and cell expansion; which process is enhanced depends on both the auxin concentration and the cell type. Endogenous auxin appears to promote stem and primary root elongation by enhancing responsiveness to GA (61). Endogenous auxin indirectly promotes dormancy of axillary buds (62), thereby reducing branching in above-ground tissue. However, application of exogenous auxin to intact wild-type plants inhibits primary root growth while simultaneously promoting the formation and elongation of lateral roots, which are initiated from a different stem cell population than the primary root (63). Clearly, the role of auxin is highly cell-specific even in wild-type plants.

Considering first the above-ground plant organs, hypocotyls in gpa1 and agb1 mutants are shorter than in wild-type plants because they contain fewer cells (24, 64). If this effect is related to auxin, it must reflect altered auxin regulation of cell division, because assays of auxin-related hypocotyl cell elongation show no substantial differences between gpa1 and agb1 mutants versus wild-type plants (24). gpa1 leaves similarly show reduced cell division, whereas GPA1 overexpressors show ectopic cell division (30). These results imply a role for G proteins in stimulating cell division in above-ground tissues.

In roots, the G protein component mutants show clear alteration in auxin responsiveness of lateral root formation (24). In the presence of auxin, agb1 mutants and GPA1 overexpressors form many more lateral roots than do wild-type plants. In the absence of exogenous auxin, gpa1 mutants form fewer lateral roots than do wild-type plants. These results are consistent with a model whereby free βγ subunits (postulated to be more abundant in the gpa1 genotype and less abundant in the agb1 and GPA1 overexpressor genotypes) negatively regulate the auxin-induced cell division that leads to the production of lateral root primordia. According to this scenario, the status of the active heterotrimer in auxin response would be signaled not by Gα but rather by interaction of Gβγ with downstream effectors. Auxin may act in part by reducing the abundance of the negative regulator AGB1, because in wild-type seedlings, the amount of AGB1 transcript is decreased by auxin treatment while that of GPA1 is increased (24).

How Can Only Two Plant Heterotrimers Mediate So Many Plant Hormone Responses?

Even though several of the important plant hormones listed in Fig. 1 have yet to be evaluated for altered responsiveness in the G protein component mutants, and responses to the other hormones have been only incompletely studied, it is evident that G protein signaling modulates a plethora of hormonal effects, thereby influencing numerous developmental processes and physiological responses. The information summarized above suggests at least two explanations for the versatility of plant G protein signaling.

A Limited Number of Heterotrimeric Combinations Transduce Signals from Multiple Hormones

In seeds, germination responses to GA, BR, and ABA are all altered in G protein component mutants (46). Not all of these responses have been evaluated in the rice rga1 mutants, but responses to GA and BR seem similarly altered in rice seeds.

Two models have been put forth to explain such results (50). On the basis of the results observed in rice, it has been suggested that there may be a high-sensitivity GA pathway that is transduced through the rice Gα subunit RGA1 and (presumably) a GPCR, as well as a low-sensitivity pathway, mediated by a non–GPCR-type GA receptor, that functions independently of G protein signaling (Fig. 3A). Thus, in the rice d1 mutant, high sensitivity to GA is lost but low sensitivity to GA is retained. According to this model, RGA1 might actually respond to only one hormone, GA, but with lack of input from the high-sensitivity GA pathway, the ultimate effect would be an apparent increased sensitivity to signals that are inhibitory for germination.

An alternative model (46, 50) is that GA does not signal via a GPCR and G protein, but rather Gα modulates the sensitivity of the seed to GA, perhaps by increasing the effectiveness of BR in promotion of GA sensitivity (Fig. 3B). Thus, when Gα is eliminated, sensitivity to GA is reduced even though Gα does not function directly in the GA signaling pathway. Consistent with this possibility, overexpression of GPA1, while causing GA hypersensitivity, does not eliminate the GA requirement for Arabidopsis seed germination (46).

A third model would add another level of complexity, namely that GPA1 influences not only sensitivity to GA but also sensitivity to all the stimuli that regulate seed germination (Fig. 3C). In both the second and third models, the G protein does not in fact transduce many hormonal signals; rather, it acts as a modulator of many hormonal responses. Thus, G proteins might play an important role as "integrators of cell status," perhaps signaling something as general as metabolic status or "homeostasis." Global identification of proteins that physically interact with plant G protein components could help substantially in differentiating among these three models.

G protein mutants are not the only mutants to show alterations in responsiveness to several different hormones (65). Indeed, such "cross-talk" in hormone signaling may be the norm rather than the exception in plants. For example, ABA-deficient mutants have been isolated on the basis of the phenotype of germination insensitivity to the GA biosynthesis inhibitor paclobutrazol (66), whereas some ethylene-insensitive mutants show enhanced sensitivity to inhibition of seed germination by ABA (67). Mutants in a PP2A regulatory subunit show phenotypes related to auxin, ABA, and ethylene signaling (43, 68, 69). Both the second and third models can also apply to any proteins involved in hormonal cross-talk, unless such proteins function directly in hormone biosynthesis.

Limited Heterotrimeric Combinations Yield Different Outputs in Different Cell Types

From the information summarized above, it is clear that one hormone can have drastically different effects on different cell types. For example, GA opposes ABA action in seeds, but there is little evidence for GA responsiveness of guard cells; auxin inhibits lateral branching in shoots but promotes formation and growth of lateral roots. Cell-specific roles of G proteins apparently confer some of this specificity, as exemplified by the fact that certain organs of the rice Gα mutant exhibit drastically reduced GA sensitivity, whereas in other organs, sensitivity to this hormone is essentially wild-type (50).

One way to achieve cellular specificity in G protein signaling would be by cell-specific expression of GPCRs. Only two candidate GPCRs have been identified in plants to date, GCR1 and RGS1, but there may be more GPCRs that are not identifiable by sequence similarity searches. There also could be different levels of basal activation of a given GPCR in a given cell type, leading to differential susceptibility to endogenous ligands of the various classes summarized in Fig. 2. In addition, protein-protein interactions and posttranslational modifications will also affect GPCR activity status; as yet, these are essentially unexplored topics in plants. Finally, non–GPCR-type regulators of the heterotrimer are beginning to be uncovered in animals (13). These may exist in plants as well, but they would be difficult to identify by bioinformatic approaches, because proteins with this function may share neither sequence nor structural homology.

At the level of the heterotrimer, although there are only two known heterotrimeric combinations, it is possible that additional Gγ subunits will be uncovered, because sequence homology is not a strong predictor of Gγ subunits (21). There is also diversity in the ways that the heterotrimer can couple with downstream signals. In mammalian systems, some signals are transduced primarily through Gα and others through the Gβγ dimer, whereas some signal via both of these elements (1). Evidence is accumulating that such diversity also holds true for plants; for example, the modulation of auxin-regulated lateral root production appears to be mediated primarily by AGB1. Finally, patterns of expression of downstream effectors [of which only a handful have been identified in plants to date (9)] can also confer specificity of response.

Thus, it may turn out that cell specificity of G protein–based hormonal signaling in plants can be readily accounted for by combinatorial specificity among all of the proteins that make up the G protein signaling "super-pathway." Clearly, much research is required in this realm to uncover this postulated richness of G protein signaling in plants.

Why Aren’t Plant G Protein Component Knockout Mutations Lethal?

Intuitively, one would expect that a gene knockout is less likely to be lethal if there exist in the genome multiple copies of the gene, or closely related genes, or if the gene has a very restricted pattern of expression. When multiple, closely related isoforms are encoded, one expects that there may be true biochemical or "proximate" redundancy of function, whereas if gene expression is restricted, especially to a nonessential part of the plant (trichomes, for example), one expects that the deficiency may be surmountable. However, in the case of G protein component genes, the situation is exactly the opposite: In both Arabidopsis and rice, Gα and Gβ are each represented by single genes and only two Gγ genes have been identified. Moreover, these genes are widely expressed throughout the organism (7). Yet gene knockouts of plant G protein components do not lead to lethality, in contrast to some Gα mutants of yeast and mouse (70, 71) or to clear "genetic diseases" as seen in animal models (72, 73).

For a possible explanation to this puzzle, we may need not look any further than the phenomenon of hormonal cross-talk in plants, already discussed above. Because input from many different hormones (and other signals) is integrated into a given outcome, such as hypocotyl cell size or the decision to germinate, functional or ultimate redundancy may take the place of biochemical or proximate redundancy. In other words, there are probably many paths that plant cell metabolism can take to achieve the same final outcome of a particular seed status or cell size. Therefore, when G protein signaling is genetically altered in the mutants, the presence of functional redundancy provides at least a partial compensation that is sufficient to maintain plant viability. This explanation is especially likely if G proteins function according to the third model mentioned above (Fig. 3), where their input into such phenomena is not as primary, directly coupled transducers of a specific hormonal ligand, but rather as modulators of a multihormone-based response.

Phenotypic Plasticity and G Proteins: A Speculative Conclusion

From the observation that none of the effects of plant G protein component mutation as yet uncovered appear to be "life-threatening," one might reach the conclusion that G proteins do not play as central a role in plant physiology as they do in animal physiology. However, such a conclusion may be premature. The hallmark of plant growth is phenotypic plasticity, defined as the ability of any given genotype to yield different phenotypes in response to different environments (74, 75). It is argued that because plants are sessile, they have evolved increased plasticity over their mobile animal counterparts, thus allowing improved acclimation to a changing but inescapable environment (76, 77). Because G protein component mutation may eliminate inputs that contribute to the achievement of functional redundancy, such mutations may reduce the possibility that the mutants will be able to alter their phenotype and acutely acclimate to varying environmental conditions (Fig. 4A). Paradoxically, a divergence from the wild-type phenotype under any one condition could potentially signify a decrease in overall plasticity, if the mutant phenotype is "locked in" over a range of conditions in which the wild type plastically responds. One potential example is the range of stature of the rice Gα subunit mutant d1: The dwarf d1 mutant shows a much smaller range of absolute plant height than do nondwarf varieties. An extreme hypothetical example is depicted in Fig. 4A, where the function describing the relationship between phenotype and environment (the "reaction norm") of the mutant is a horizontal line, indicating a complete lack of plasticity.

Alternatively, G protein mutation could affect phenotypic variation [here defined as the range of phenotypes observed within any one environment, and sometimes referred to as "developmental noise" (74)] but not the amount of plasticity (Fig. 4, B and C). Whether a reduction in variation would ultimately be beneficial or detrimental would depend on whether the mutant phenotype converged on (Fig. 4B) or diverged from (Fig. 4C) the "ideal" phenotype. A third possibility is that G protein mutation does not affect the amount of either plasticity or phenotypic variation, but simply shifts the reaction norm (Fig. 4D). Again, whether such a shift would be helpful or harmful would depend on whether the shift is toward or away from some ideal phenotype.

In the absence of data to address the hypotheses of Fig. 4, why speculate that plant G proteins may be important for plasticity? The reason is inherent in the extremely pleiotropic nature of the G protein mutations themselves. With G proteins affecting so many diverse plant developmental and physiological responses, it seems reasonable to hypothesize that plasticity under a range of environmental conditions may be affected, with possible attendant consequences for plant fitness.

Indeed, recent plant ecophysiological studies on other Arabidopsis mutants are revealing that the extent of plasticity of a given genotype is an important component of the fitness equation. For example, Arabidopsis mutants that lack photoreceptors for perception of specific light signals are at a disadvantage in certain environments (78). Therefore, it will be of interest to evaluate the plasticity, and ultimately the reproductive success, of G protein mutants under a range of environments, including nonoptimal conditions. Such experiments may reveal vital contributions of plant G proteins to fitness after all.

Given the rate of anthropogenically based changes in the global environment, plasticity may be one of the more important traits for plant species survival. The extent of plasticity of human phenotypic response, although much more limited than that of plants, has important implications for human health (79). Thus, despite their scarcity in "green genomes," studies on plant G proteins and their regulators may prove informative to a wide variety of disciplines.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
  65. 65.
  66. 66.
  67. 67.
  68. 68.
  69. 69.
  70. 70.
  71. 71.
  72. 72.
  73. 73.
  74. 74.
  75. 75.
  76. 76.
  77. 77.
  78. 78.
  79. 79.
  80. 80.
  81. 81.
View Abstract

Stay Connected to Science Signaling

Navigate This Article