Evidence Mounts for Receptor-Independent Activation of Heterotrimeric G Proteins Normally in Vivo: Positioning of the Mitotic Spindle in C. Elegans

See allHide authors and affiliations

Science's STKE  19 Aug 2003:
Vol. 2003, Issue 196, pp. pe35
DOI: 10.1126/stke.2003.196.pe35


Examples of the activation of heterotrimeric G proteins in vivo by any means other than through activated cell surface receptors have been limited to pathophysiological phenomena. With the discovery of proteins apart from receptors that facilitate guanine nucleotide exchange and affect G protein subunit dissociation directly, however, the notion of receptor-independent modes of activation in normal circumstances has become a subject of great interest. Three recent publications, each focusing on G protein regulators (GPRs) in asymmetric positioning of the mitotic spindle in the early Caenorhabditis elegans embryo, provide substantial support for the likelihood of such a form of activation. The C. elegans proteins GPR-1 and GPR-2 each contain a G protein regulatory motif, which supports interaction with Gαi-like subunits. Inactivation of the genes encoding GPR-1 and GPR-2 prevents the correct positioning of the mitotic spindle in the one- and two-cell embryo. This phenotype is identical to that achieved by inactivation of genes encoding the Gα subunits GOA-1 and GPA-16. Because signaling in the one- and two-cell embryos is "intrinsic," the data suggest a GPR-dependent, receptor-independent mode of G protein activation. The GPRs interact preferentially with the guanosine diphosphate (GDP)-bound form of Gα subunits, and the GPR motif per se exhibits GDP dissociation inhibitor activity. The actions of the GPRs imply that GDP•Gα•GPR is a key intermediate or effector in force generation relevant to mitotic spindle positioning.

The long-standing assumption that, in a normal physiological context, heterotrimeric guanine nucleotide-binding proteins (G proteins) are activated through receptor-mediated processes alone has been subject to increasing challenge. For example, a recently identified activator of G protein signaling, AGS1, activates G proteins through a mechanism overtly similar to that used by classical seven-transmembrane domain (7TM) receptors, promoting guanosine diphosphate (GDP)/guanosine triphosphate (GTP) exchange (1); however, AGS1 does not resemble 7TM receptors, nor is it presumed to be present at the cell surface. Other AGSs, AGS2 and AGS3, activate one or more G proteins by interacting with Gβγ or Gα directly to effect the requisite subunit dissociation. The difficulty in asserting receptor-independent modes of G protein activation, nevertheless, has been in proving that these modes of activation are in fact normally operant in the intact cell, and in identifying the precise set of circumstances in which they are engaged. Some of the most provocative data, at least in one area of cell function, are now emerging at the level of mitotic spindle positioning in metazoans. Recently published papers by Srinivasan et al. (2), Colombo et al. (3), and Gotta et al. (4), all focusing on the roles of so-called G protein regulators (GPRs) in asymmetric positioning of the spindle in the early C. elegans embryo, specifically address the topic of receptor-independent G protein function.

Asymmetric positioning of the spindle, in conjunction with polarity cues, ensures the correct segregation of cell fate determinants between daughter cells. In the one-cell C. elegans embryo, the spindle is positioned along the anterior-posterior cell axis and is displaced toward the posterior during anaphase. Displacement of the spindle is achieved through unequal anterior-posterior forces exerted through astral microtubules that terminate at the cell cortical membrane (5). The positioning of the spindle causes asymmetric cell division transverse to the anterior-posterior axis, yielding a larger, anterior AB cell and a smaller, posterior P1 cell. Gene inactivation studies reveal that G protein subunits play a critical role in positioning the mitotic spindle (6, 7). Inactivation by RNA interference (RNAi) of the genes encoding the Gαi-like subunit GOA-1 and the structurally similar subunit GPA-16 (referred to as GOA-1/GPA-16 when the effects of each are presumed interchangeable) causes multiple defects, including, in the one-cell embryo, symmetric positioning of the spindle.

All three studies focus on the products of the predicted genes F22B7.13 and C38C10.4. The two products are closely related proteins that each contain two tetratricopeptide motifs and one GPR [or GoLoco (8)] motif. The GPR motif is common to certain proteins in mammalian systems that interact with Gαi subunits, for example, AGS3 and a few RGSs (regulators of G protein signaling) (8, 9). On the basis of these sequence similarities, the gene products are assigned the names GPR-1 and GPR-2 (referred to collectively as GPR-1/GPR-2, or GPRs). Inactivation of gpr-1 and gpr-2 together by RNAi was found to result in only limited spindle movement in the one-cell embryo, resulting in AB and P1 cells of equal size, and also aberrant spindle positioning in the two-cell embryo. These and other events following gpr-1 and gpr-2 inactivation are essentially identical to those following goa-1 and gpa-16 inactivation. Thus, GPR-1/GPR-2 and GOA-1/GPA-16 appear to be working in tandem or in parallel in spindle positioning. Of considerable interest, the peak velocity of the posterior spindle pole in wild-type cells, as determined by Colombo et al. after spindle severing in vivo, was 40% greater than that of the anterior one, with the posterior pole traveling farther; however, when gpr-1 and gpr-2 were inactivated, both poles exhibited the same, and considerably reduced, peak velocity. These data implicate both GPR-1/GPR-2 and GOA-1/GPA-16 in force generation relevant to spindle positioning.

How, specifically, are GPR-1/GPR-2 and GOA-1/GPA-16 arranged to effect mitotic signaling, and what do any of the observations have to do with receptor-independent signaling? GOA-1, by immunofluorescence, is present at the cell cortex and, to a lesser extent and diffusely, around the spindle asters. GPR-1 is also present at the cortex and spindle asters. GPR-1/GPR-2 aside for the moment, one might envision that GOA-1 and GPA-16 at the membrane are subject to activation by agonists through 7TM receptors, such that they could modulate levels of second messengers relevant to force generation. Asymmetry in force would be achieved through spatial gradients in agonist concentrations or through asymmetric positioning of receptors, GOA-1/GPA-16, proteins that stabilize or degrade second messengers, or relevant effectors.

One of the arguments against activation of GOA-1 or GPA-16 through 7TM receptors, however, is the autonomous, or intrinsic, nature of mitotic spindle positioning in one- and also two-cell embryos. The chitin eggshell provides a barrier to agonists, and spindle orientation in P1 cells is not influenced by cell-cell contact. Yet the argument, although strong, is not airtight. The generation of autocrine factors, soluble or membrane-bound, that serve as agonists for 7TM receptors cannot be precluded. Such factors may be produced in a spatially defined manner or may participate in receptor-mediated processes otherwise asymmetric in nature. Nor can one preclude the relevance of "constitutive" (agonist-independent) activity of certain 7TM receptors that act in parallel to regulated processes. Smoothened is an example of a 7TM protein (although not unequivocally a G protein-coupled receptor) that favors an active conformation in the absence of any ligand (10). A relative, Frizzled, is known to participate in extrinsic control of spindle positioning in the C. elegans embryo at the four-cell stage (11).

Yet in no instance has a requirement for anything similar to GPR-1 and GPR-2 in 7TM receptor signaling been noted. This fact alone provides a strong argument for a unique signaling paradigm, as proposed by all three groups. The GPR proteins are disposed predominantly to the posterior aspect of the cell by anaphase in the one-cell embryo (3, 4), consistent with asymmetric force generation. The presence of the GPR motif is key to speculation regarding function. In mammals, proteins containing this motif interact, like Gβγ, with the GDP-bound form of Gαi-like subunits, compete with Gβγ for binding to this form of subunit, and inhibit dissociation of GDP from Gαi [that is, they are GDP dissociation inhibitors (GDIs)] (12, 13). The interaction of GPR-1 with GDP-bound as opposed to GTP-bound GAO-1 is shown in all three papers, and Gotta et al. additionally demonstrate that the GPR motif in GPR-1 has GDI activity. One might consider at the outset the possibility that GPR-1/GPR-2 competes with Gβγ for the Gα subunit, and that the released Gβγ is responsible for triggering asymmetric force generation. Release of Gβγ is, in fact, the activating activity originally defined for AGS3. However, although Gβγ is involved in certain aspects of spindle orientation, the phenotypes observed when genes for Gβ or Gγ are inactivated in the C. elegans embryo are quite distinct from those observed when genes for GPR-1/GPR-2 and GOA-1/GPA-16 are inactivated (7). It is therefore Gα, not Gβγ, that is implicated in force generation in C. elegans. A role for a Gαi-like subunit coupled to another GPR domain-containing protein, PINS (Partner of inscutable), in asymmetric spindle positioning in Drosophila neuroblasts has also emerged (14).

What, then, is the form of Gα that signals? The simplest explanation is that the GDP-bound forms of GOA-1 and GPA-16, not the GTP-bound forms, mediate spindle positioning (Fig. 1). In this instance, GPRs serve as upstream regulators and GDP•Gα is the entity capable of engaging with proteins relevant to force generation. An attractive variant of this scheme is that GPRs are both the upstream regulators and components of the downstream effector; specifically, GPRs compete with Gβγ for GDP•Gα, and the GPR•Gα complex as a whole is relevant to force generation. The idea of GDP•Gα as a signaling entity has, however, scant precedent. As discussed by Gotta et al., the idea requires a yet-undefined mechanism by which Gβγ might eventually recouple with Gα and thereby terminate signaling (but see below). And, although perhaps a question of semantics, one might question whether a GDP•Gα signaling entity represents an "activated" form of Gα in any classical sense, and therefore whether the process should be proposed as a receptor-independent means of activation.

Fig. 1.

Models of receptor-independent modes of G protein activation in force transduction relevant to mitotic spindle positioning in C. elegans. Three models of G protein activation are consistent with the data supporting the role of GPR-1/GPR-2 in the regulation of GOA-1/GPA-16. GPR-1 or GPR-2 competes in all three instances with Gβγ for the G protein α subunit, retaining the Gα subunit in a GDP-bound form, but in one case supports subsequent exchange for GTP. (A) GDP•Gα, either alone or as presented by GPR-1 or GPR-2, is recognized by an effector relevant to force transduction. (B) The GDP•Gα•GPR complex, as a whole, is recognized by the effector. (C) The GDP•Gα•GPR complex is recognized by RIC-8 [or some other guanine nucleotide exchange factor (GEF)] in much the same way that GDP•Gα•Gβγ is recognized by 7TM receptors, with the result that RIC-8 is able to facilitate exchange of GDP for GTP. It is GTP•Gα, free from GPR and RIC-8 if the analogy to Gβγ and 7TM receptors holds, that interacts with the effector.

Yet another possibility is based on the documented ability of Gβγ to facilitate communication of Gα with 7TM receptors (15), which sets the stage for agonist-effected activation of the G protein. Gβγ stabilizes G protein interactions with receptors in part by stabilizing the GDP conformation of Gα, which 7TM receptors favor, and in part by interacting directly with both Gα and receptor. The similarity of GPRs to Gβγ with regard to GDI activity raises the question of whether GPRs also position Gα for activation by receptors or some other form of guanine nucleotide exchange factor. The question of receptor has yet to be addressed, but both Srinivasan et al. and Gotta et al. propose that GPRs could place GAO-1 and GPA-16 into a position of susceptibility to a recently identified guanine nucleotide exchange factor, RIC-8 (16). Ric-8 is required for correct embryonic spindle positioning in C. elegans, and loss-of-function phenotypes for ric-8 resemble those of gpr-1/gpr-2 (17). RIC-8, in this model, would recognize not Gαβγ but instead the Gα•GPR complex, and subsequently would effect exchange of GDP for GTP on the Gα subunit. If the analogy to Gβγ-like function holds, the GTP•Gα subunit would then dissociate from GPR-1/GPR-2 to engage effectors relevant to force generation. Signal termination would be achieved by hydrolysis of the bound GTP.

An obvious question, given the topic of receptor-independent modes of G protein activation, is the identity of signals that cause GPR-1 and GPR-2 to interact with GOA-1 and GPA-16. If not agonists in the traditional sense, then what? The signals themselves will likely emerge with a more complete understanding of the order and nature of linkages among polarity factors, LIN-5 (2, 4), the GPRs, and the G protein subunits. Possibilities for mechanisms include anything that might decrease the stability of the interaction between the Gα subunits and Gβγ (such that GPRs can gain access to the Gα subunits more readily) or anything constituting an "activation" of the GPRs. With regard to the latter, Blumer et al. (18) have recently determined that the region of AGS3 containing several GPR motifs is a substrate for phosphorylation by the serine-threonine kinase LKB1 and that introduction of a phosphate into a GPR motif reduces the ability of the motif to interact with Gα. Of interest, the LKB1 ortholog in C. elegans is PAR4, a polarity factor (although not asymmetrically distributed) (19). Perhaps, then, the GPRs are phosphorylated constitutively, and activation and consequent interaction with Gα subunits is achieved through activation of a phosphatase. Phosphorylation, on the other hand and relating more to a process of deactivation, may represent a mechanism by which an already existing interaction of a Gα subunit with a GPR is terminated to allow recoupling of the subunit to Gβγ. Activation of a GPR might be achieved by yet undefined protein-protein interactions. Other questions include, of course, the identity of effectors engaged by GDP•Gα, GDP•Gα•GPR, or GTP•Gα.

Proteins that use GPR motifs, or that are involved in asymmetric spindle positioning, are by no means the first or last word in receptor-independent forms of G protein activation. Precedents for receptor-independent modes of G protein activation have existed for some time in pathophysiological contexts. Gs can be activated through adenosine diphosphate (ADP) ribosylation by cholera toxin; Gs and Gi can also be activated through certain somatic mutations (20). The activation in both instances is due to an inhibition of the GTPase activity intrinsic to the Gα subunit, therefore mimicking in an indirect and sustained fashion the actions of receptors. Several G proteins, for example Gz and G12, are subject to phosphorylation, which inhibits the interaction of the Gα subunits with Gβγ (21, 22). Phosphorylation of Gα may increase the proportion of the subunit bound to GTP, depending on the subunit's relative affinities for GTP and GDP or any other factors (AGSs?) that might better recognize and activate the monomeric subunit. Last, sphingosine 1-phosphate (S1P) has been shown to regulate stomatal apertures and guard cell ion channel activities in a G protein-dependent fashion in Arabidopsis thaliana (23). The absence of an S1P-like receptor suggests an activation of the G protein directly or through proteins other than a receptor.

With the demonstration that GPR-1 and GPR-2, each containing a motif supporting interaction with Gαi-like subunits in C. elegans, are required in a Gα-dependent generation of force relevant to spindle positioning in the one- and two-cell embryo, the idea of a 7TM receptor-independent mode of G protein activation in a normal, physiological context is strengthened. An even more compelling argument will emerge with the elucidation of how the interactions between GPRs and Gα subunits are in fact regulated in the context of cell division. The extension to other aspects of cell function in C. elegans and vertebrates, of course, will prove important.


  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.
View Abstract

Stay Connected to Science Signaling

Navigate This Article