There Are GAPS and There Are GAPS

See allHide authors and affiliations

Science  03 Jan 1997:
Vol. 275, Issue 5296, pp. 42-43
DOI: 10.1126/science.275.5296.42

Although their main function is to regulate other proteins, guanine nucleotide-binding proteins (G proteins) are also guanosine triphosphatases (GTPases), cleaving guanosine triphosphate (GTP) to form guanosine diphosphate (GDP). Because of this activity, they oscillate between GTP- and GDP-bound states, and thus regulate diverse processes such as protein synthesis, cytoskeleton assembly, vesicle transport, and signal transduction. The superfamily comprises both small monomeric and large multimeric G proteins, but for all members, the release of bound GDP and the binding of GTP are highly regulated processes (1). The GTPase activity of small G proteins, such as EF-Tu and Ras, is stimulated by associated proteins called GAPs (GTPase activating proteins) (2). But GAPs for most large G proteins had not been described until recent work identified members of the regulators of G protein-signaling (RGS) family as GAPs for this subfamily.

The large heterotrimeric G proteins involved in signal transduction have αsubunits, which are related to small G proteins, and βγsubunits that exist as a single complex. Both Gα and Gβγ can independently transmit signals (3). Signal termination for both Gα and Gβγ subunits likely occurs through GTP hydrolysis. How the GTPase terminates signaling through Gα subunits is easily understood given the observation that GDP-Gα subunits have much lower affinities for effectors than do GTP-Gα complexes (4). But how signaling through Gβγ is terminated has not been as clear. A report in Cell by Gilman and co-workers (5) and two others in Nature (6) identifying two members of the RGS family, GAIP and RGS4, as GAPs for members of the Gαi family and other recent papers shed light on this issue.

A four-component heterotrimeric G protein-signaling system

The resting G protein is an αβγ heterotrimer with GDP bound to it. Activated receptor promotes the release of GDP, the binding of GTP, and dissociation of GTP-Gα from the Gβγ complex. GTP-Gα and Gβγ can now interact with their effectors and propagate the signal. RGS stimulates (+) the GTPase activity of the Gα subunit, resulting in the accumulation of GDP-Gα, which re-forms the stable heterotrimer. Free Gβγ leads to dissociation of Gβγ from effector, thus terminating signal propagation. R, receptor; E, effector; αβγ, the heterotrimeric G protein; and RGS (GAP), stimulator of the GTPase of the Gα subunit.

Members of the RGS family have been identified in yeast, Caenorhabditis elegans, and mammals (7). Sst2p in yeast and EGL-10 in C. elegans, homologs of RGS, suppress signal transmission by acting on the G protein-α subunit (8). Mammalian RGS can substitute for yeast Sst2p in regulating pheromone signaling (9), which is transmitted through Gβγ subunits (10). Taken together, these data suggest that RGS can regulate signaling through Gβγ subunits by modulating the activity of the GTPase of the α subunit. How would such regulation work? The positive cooperativity between Gβγ and GDP interaction with Gα subunits (11) indicates that Gβγ has a higher affinity for GDP-Gα than for GTP-Gα. If Gβγ also has a higher affinity for GDP-Gα than for effectors such as K+ channels, adenylyl cyclase, and phospholipase C, then RGS, by stimulating the GTPase catalytic activity and the accumulation of GDP-Gα could shift the steady-state level toward the inactive heterotrimeric state. Alternatively, if the affinity of Gβγ for GDP-Gα and effectors is in the same range, a much greater concentration of Gα in comparison to the effector could push Gβγ to the heterotrimeric state. Either way, RGS would be essential to build up the concentration of GDP-Gα after activation of the G protein, and the relative stoichiometry of effectors to Gα is likely to be important for Gβγ-mediated signaling.

This mechanism may be operative for Gi. Here the Gα subunit would define the specificity of receptor interaction and in conjunction with RGS define the lifetime of the active Gβγ complex. Different Gαi isoforms have different rates of GDP release (12), and if the different RGS isoforms stimulate the GTPase rates to different extents, then a wide range of timing can easily be obtained. Such facile temporal regulation would be particularly valuable for Gβγ regulation of K+ and Ca2+ channels. RGS stimulation of Gαi1 GTPase could explain the discrepancy between the 20-fold faster deactivation rate of the muscarinic K+ channel upon removal of the agonist, as compared to the basal GTP hydrolysis rate of purified Gi (13). The type of regulation described above presumes that activation results in subunit dissociation and that all of the free Gβγ can regulate effector, although this has not been proven.

Experiments in E. Ross's laboratory have shown that Gβg inhibits GAP stimulation of the GTPase-Gα (14), suggesting that Gβγ and RGS may interact with overlapping regions of Gα. Additionally, the RGS (GAP) proteins interact with GDP-Gα with much lower affinity than they do with the transition-state complex (15). These properties would facilitate GTP hydrolysis on Gα followed by dissociation of RGS from Gα and the reassociation of Gα and Gβγ, further supporting the notion that RGS regulated GTPase is the major mechanism for turning off Gβg signaling. Clearly, for Gi/Go coupled systems we have moved from a three- to a four-component system (see figure).

In contrast, GAPs for monomeric G proteins function as On-Off switches because the basal GTPase activity of the monomeric G proteins is very low. RGS is poised to be an effective modulator of transmembrane signaling through G protein pathways and can serve as a putative locus for interactions between signaling pathways.

Comparison of the work by Gilman and co-workers (5, 15) with that from the Wittinghofer group (6, 17) suggests that GAPs regulate the activity of small and large G proteins by different mechanisms. GDP-Gα can bind AlF4—small G proteins, such as Ras, cannot, but GDP-Ras protein can bind AlF4 when associated with its GAP protein (16). Crystallographic studies on Gα subunits indicate that the AlF4·GDP-Gα complex represents the transition state of the SN2 reaction that occurs during GTP hydrolysis (18). Thus for Ras, GAP is thought to contribute residues crucial for formation of the transition state (16). Prominent in this context is an Arg (Arg178 in Gαi1) present in Gα subunits but not in Ras or other small G proteins. The crystal structure of the active domain of the p120GAP and a manual docking model of Ras and the GAP active domain suggest that an Arg from GAP could be used for GTP hydrolysis by Ras (17). Thus, for small G proteins the main function of GAP would be to move the Ras-GTP complex to a transition state for nucleotide hydrolysis. It is unlikely that this is the primary mechanism by which RGS stimulates the GTPase activity of Gα, since Gα by itself can form a transition-state complex (18). However, RGS has the highest affinity for AlF4·GDP-Gα (15) indicating that it preferentially binds to the transition-state complex and thus promotes hydrolysis. The precise mechanism by which RGS promotes GTP hydrolysis by Gα remains to be determined.

Just when we thought that the basic G protein-signaling system had been well defined, nature has provided us with a surprise. There are probably more surprises to come.


  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.
View Abstract

Stay Connected to Science

Navigate This Article