G-diversity discovered


Studies on heterotrimeric G-proteins have played a major role in clarifying how hormonal and environmental signals are converted into physiological responses in cells. The Nobel Prize in Medicine in 1994 was awarded to Alfred Gilman and Martin Rodbell for their pioneering work on these proteins, which they carried out primarily in mammalian cells. Since then it has become clear that the basic components of the heterotrimeric G-protein complex – Gα, Gβ and Gγ– are broadly similar among all the metazoa both in structure and in function (Wettschureck & Offermanns, 2005). Early steps in the G-protein signaling pathway include the ligand (e.g. hormone) or environmental activation of a G-protein-coupled receptor (GPCR), typically in the plasma membrane, which then binds to Gα in the membrane-bound trimeric complex and acts as a guanine nucleotide exchange factor (GEF), catalyzing the exchange of GTP for the GDP bound to Gα. This ‘activates’ the G-protein and triggers a series of transduction steps that begin with the separation of Gα from the Gβγ dimer, each of which then becomes an independent signal amplifier by binding to and altering the activity of downstream enzymes or ‘effectors’ (Fig. 1). The activity cycle is terminated when Gα, which is a GTPase, converts its own bound GTP to GDP, a reaction typically enhanced by a GTPase-activating protein (GAP). After this, Gα re-associates with the Gβγ dimer and reforms an inactive heterotrimeric complex, poised again to be turned on by an activated GPCR (Neves et al., 2002). This basic pattern of G-protein signaling (Fig. 1) is found throughout the metazoa. Although the heterotrimeric G-protein complex is now well documented in plants, recent data raise doubt about how closely the activity of this complex in plants mimics that in animals (Chen, 2008). In this issue of New Phytologist, Bisht et al. (pp. 35–48) provide valuable new data that help address this issue.

Figure 1.

 A schematic illustrating the main steps in G-protein signaling as typically observed in metazoans. The guanine nucleotide exchange factor (GEF) activity of a ligand-activated G-protein-coupled receptor (GPCR) removes GDP from Gα and exchanges it for GTP. This activates the G-protein, resulting in the separation of Gα from the Gβγ dimer, each of which can then induce activity changes in downstream effectors. New data in Bisht et al. (this issue of New Phytologist, pp. 35–48) reveal that different Gα proteins in soybean differ in their GTPase activities, and this has implications for how G-protein signaling in plants could either resemble or differ from that in animals. (Reproduced from Chen (2008), with modification.)

‘… it is now clear that G-protein signaling in plants is not as different from animal systems as previously believed.’

Humans and most metazoans have multiple versions of Gα, Gβ and Gγ that can mix and match, allowing for numerous different heterotrimeric complexes; for example, there are more than 1000 potential combinations in humans (Wettschureck & Offermanns, 2005). Humans also have almost 1000 different GPCRs, all characterized by a globular shape and seven-transmembrane segments (Bjarnadóttir et al., 2006). In many metazoans these receptors are more numerous and diverse than any other type of membrane receptor. However, initial studies in plants have suggested that the number and diversity of heterotrimeric G-proteins and of GPCRs are much lower. Both Arabidopsis and rice (Oryza sativa) have only one classic Gα, one Gβ and two Gγ proteins, and their genomes encode only a few proteins that have the seven-transmembrane structure and other features similar to metazoan GPCRs (Perfus-Barbeoch et al., 2004). Another major difference discerned in early studies of plant heterotrimeric G-proteins was that Gα appeared to have a relatively low GTPase activity and a high spontaneous GEF activity (Johnston et al., 2007), so that under steady-state conditions, according to theoretical calculations, c. 99% of the Gα in Arabidopsis would be GTP bound, whereas in mammals only c. 10% would be predicted to be GTP bound (Chen, 2008). Based on this finding, Johnston et al. (2007) proposed that the rate of the GTPase activity of Gα, rather than the GEF rate of a GPCR, would be the limiting step in the activity cycle of Gα in plants. Consistent with this hypothesis, the GEF activity of candidate GPCRs characterized thus far in plants is quite low. Taken together, the early findings on heterotrimeric G-proteins in Arabidopsis and rice favored the opinion that in plants the Gαβγ combinations are far less diverse and act more independently of GPCRs than is the case in animals.

This opinion will be changed by the important findings reported by Bisht et al. These authors investigated the diversity of genes encoding Gα, Gβ and Gγ proteins in the palaeopolyploid genome of soybean (Glycine max), a species thought to have formed by the joining of two ancient parental genomes. By using a polyploidy species, Bisht et al. increased the likelihood of discovering multiple genes coding for each of the three different subunits. As might be expected, then, the authors did identify multiple versions of Gα (four) and Gβ (four), although the number of Gγ proteins they found (two) was surprisingly the same number as in Arabidopsis and rice. They carried out interaction assays in yeast that demonstrated specific interactions among the three subunits, and they expressed yellow fluorescent protein (YFP) hybrids of the Gα proteins in tobacco to show that they were localized on the plasma membrane. These data revealed the most extensive network of heterotrimeric G-proteins yet found in plants.

Bisht et al. then heterologously expressed and purified the recombinant proteins and characterized their GTP-binding and GTP-hydrolysis activities to establish them as authentic G-proteins. These experiments produced the key and novel finding that, although the GTP-binding activity of all four Gα proteins was similar, their GTPase activities were significantly different. GmGα1 and GmGα4 exhibited a slow rate of GTP hydrolysis, similar to that of the Arabidopsis Gα, AtGPA1. GmGα2 and GmGα3, however, had significantly higher GTPase activities, more comparable with that of metazoan Gα proteins. Because these two GmGα proteins have a fast GTPase activity, it is now clear that G-protein signaling in plants is not as different from animal systems as previously believed. One caveat here is that all the activity data are obtained from recombinant proteins. Their reliability depends on the assumption that the activities of the heterologously expressed Gα will not differ from those in plant cells, which could be modified post-translationally in ways that would not occur during bacterial expression.

The diversity of heterotrimeric G-protein subunits in soybean is not so different from that in Drosophila, which has six Gα, three Gβ and two Gγ subunits (Katanayeva et al., 2010). True, the four Gα genes in soybean are fewer than typically found in most metazoans, but in considering this difference, it is important to note that the authors only counted the soybean genes coding for the canonical Gα subunits. There may be variants that could function in a heterotrimeric complex but that would not be readily identified by sequence analysis. There could also be splice variants, and, in fact, the authors allude to data that suggest a possible splice variant of Gα4. Moreover, soybean, like Arabidopsis, could have one or more extra-large G proteins (XLGs) that have significant sequence similarity to the Gα subunit in their C-terminal regions. The XLG2 member of this family in Arabidopsis can interact with the Gβ subunit of heterotrimeric G proteins and, in this noncanonical complex, function in defense signaling pathways (Zhu et al., 2009).

Further studies should reveal the structural basis for the difference in GTPase activity between the low level in GmGα1/GmGα4 and the high level in GmGα2/GmGα3. The functional implications of these differences for signaling mechanisms have yet to be explored, and discovering this will contribute significant depth to our understanding of the diverse ways in which G-proteins manifest their effects.

Because the sequence of the soybean genome was recently published (Schmutz et al., 2010), it will now be possible to evaluate whether the increased diversity of G-protein subunits in this species is paralleled by an increased number of GPCR candidates, and, if so, to test whether any of these candidates have significant GEF activity. Currently, the absence of a characterized plant GPCR with significant GEF activity remains a major difference between the way that plants and animals use heterotrimeric proteins to amplify input stimuli. However, future investigations could reveal examples of metazoan-like GPCR–Gα pairings where the GEF activity of the GPCR would be a major determinant of the GDP–GTP exchange that activates Gα.

Current estimates are that > 50% of plant species are, like soybean, polyploid. Thus, there is a high probability that the pioneering discoveries of Bisht et al., showing both multiplicity and functional diversity of heterotrimeric G-protein subunits in soybean, will be replicated in other species. This should reinforce the conviction that the functions of plant heterotrimeric G-proteins, although clearly unique in some respects, are not fundamentally different from the metazoan mode.