Signal transduction

New exchange, new target

Article metrics

The Ras superfamily of small guanine-nucleotide-binding (G) proteins controls many aspects of cellular behaviour, from proliferation and survival to morphology and membrane transport. Despite their varied effects, these proteins are regulated by a common mechanism — the opposing effects of stimulatory guanine-nucleotide-exchange factors (GEFs), which promote the uptake of fresh GTP to form an active, GTP-bound conformation, and inhibitory guanosine triphosphatase activating proteins, which stimulate hydrolysis of bound GTP to form the inactive, GDP-bound conformation.

On page 474 of this issue, de Rooij et al.1 report a new twist to the mechanism for control of the Ras-related protein Rap1. Their study shows that regulation of proteins in the Ras superfamily is considerably more varied than previously thought, and establishes, for the first time, the molecular details of how a Ras-related protein is controlled. As well as all this, the authors identify a new class of target protein for cyclic AMP, Earl Sutherland's original second messenger molecule, raising the possibility that several of cAMP's effects on cells may not be mediated by its previously characterized targets.

When cells are treated with a growth factor, Ras is activated by its GEF, Sos. A complex forms containing the growth-factor receptor (an autophosphorylated tyrosine kinase), adaptor proteins with SH2 and SH3 (src-homology) domains, and Sos2. Assembly of this complex is not thought to stimulate Sos directly; instead, translocation of Sos from a cytosolic location to the plasma membrane — where both the receptor and Ras are localized — is believed to give Sos improved access to Ras (Fig. 1a). However, there has been no totally convincing proof or in vitro reconstitution of this model.

Figure 1: Two types of regulatory system for Ras proteins.

a,The classic Sos pathway. Activation of a growth-factor receptor tyrosine kinase leads to formation of complexes containing adaptor proteins (Shc and Grb2) and Sos (a guanine-nucleotide-exchange factor; GEF), at the plasma membrane where Ras is also located. Increased local concentration of Sos results in nucleotide exchange on Ras to form the active GTP-bound conformation, which stimulates several effector enzymes including Raf. b,Increases in the concentration of cyclic AMP, caused in vivo by hormone activation of Gs-coupled receptors, results in the binding of cAMP to Epac, a Rap1-specific GEF identified by de Rooij et al.11. This causes a conformational change leading to increased exchange activity towards Rap1, which is localized on intracellular membranes. GTP-bound Rap1 then activates effectors that may include B-Raf.

It is also clear that other Ras GEFs exist, and that they are regulated quite differently from Sos. One, for example, the Ras guanine-nucleotide-releasing factor (Ras-GRF), is stimulated by the direct binding of Ca2+-calmodulin3 and also by the action of heterotrimeric G proteins4. Earlier this year another GEF, Ras guanine-nucleotide-releasing protein (RasGRP), was identified in the brain5. RasGRP contains Ca2+-binding motifs and a diacylglycerol-binding domain, and it is activated on treating cells with diacylglycerol — possibly because it is then relocalized to the plasma membrane.

Rap1 is a close and, until fairly recently, neglected relative of Ras. It was identified in a screen6 for molecules that inhibit the transforming function of Ras. Although Rap1 can interact with the same effectors as Ras, it is localized to intracellular membranes — this may explain why it can antagonize Ras when overexpressed. There is conflicting evidence as to whether its ability to antagonize Ras is physiologically relevant (reviewed in ref. 7), although the balance seems to be shifting away from a physiological role8. There are also data to support an alternative model for Rap1 function in which it has similar effects to Ras, particularly the ability to activate B-Raf and, hence, the mitogen-activated protein (MAP) kinase cascade9. But, in most systems Rap1 has very different effects to Ras so, although they may share some signalling pathways, Rap1 probably activates these systems in distinct ways and may also interact with unique effectors7.

Rap1 is activated in response to a range of stimuli through a number of second messenger molecules, including cAMP, Ca2+ and diacylglycerol8,10,11. The only previously identified GEF for Rap1 is C3G, which seems to function analogously to Sos. But it is unlikely that C3G mediates activation of Rap1 by cAMP, and de Rooij et al.1 now show that another GEF is responsible.

Protein kinase A (PKA) is activated by cAMP, so we could imagine a mechanism whereby cAMP activates PKA which, in turn, leads to activation of Rap1. But, by using inhibitors and mutants, de Rooij et al. found that PKA is not involved in the cAMP-induced activation of Rap1. Indeed, although PKA phosphorylates Rap1 near its carboxy terminus10, this phosphorylation is not required for cAMP-dependent activation. This led the authors to search sequence databases for possible Rap1 GEFs that might be directly regulated by cAMP, and they pulled out a protein which they named Epac (exchange protein directly activated by cAMP). As well as containing homologies to other GEFs for Ras-like proteins, Epac possesses sequences related to the regulatory subunit of PKA. De Rooij and colleagues found that Epac can bind directly to cAMP, and that its activity towards Rap1 is stimulated if the levels of cAMP in the cell are increased (Fig. 1b). Epac is, therefore, not only another GEF for Rap1, but an example of a new type of cAMP sensor protein. It is only the third such type identified to date, after cAMP-dependent protein kinases and cyclic-nucleotide-gated ion channels.

Epac seems to fit into the category of second-messenger-stimulated GEFs (such as Ras-GRF and RasGRP), rather than the adaptor-regulated GEFs like Sos and C3G. However, this new work goes well beyond any previously done on Ras-family GEFs because de Rooij et al. used purified components to show, in vitro, that the activity of Epac towards Rap1 is allosterically stimulated on binding cAMP. They also deleted the cAMP-binding domain of Epac and found that this led to activation of Epac in vitro. This indicates that the cAMP-binding domain normally inhibits the exchange activity of Epac until cAMP binds, most likely causing a conformational change that relieves the inhibition.

This is the first time that the regulation of a Ras-family GEF has been reconstituted with defined components, and de Rooij et al. provide mechanistic detail at the molecular level that is still lacking for much more intensively studied GEFs such as Sos. After years spent languishing in the shadow of its flamboyant cousin, Rap1 has finally stepped into the limelight itself. Although there may still be disagreement as to the downstream effects of Rap1 on cell function, it is clearly a broadly used sensor for second messengers such as cAMP, Ca2+ and diacylglycerol. Moreover, the localization of Rap1 on intracellular membranes, which are readily accessible to soluble second messengers, may provide further clues to its function.


  1. 1

    de Rooij, J. et al. Nature 396, 474–477 (1998).

  2. 2

    McCormick, F. Nature 363, 15–16 (1993).

  3. 3

    Farnsworth, C. L. et al. Nature 376, 524–527 (1995).

  4. 4

    Mattingly, R. R. & Macara, I. G. Nature 382, 268–272 (1996).

  5. 5

    Ebinu, J. O. et al. Science 280, 1082–1086 (1998).

  6. 6

    Kitayama, H., Sugimoto, Y., Matsuzaki, T., Ikawa, Y. & Noda, M. Cell 56, 77–84 (1989).

  7. 7

    Bos, J. L. EMBO J. 17, 6776–6782 (1998).

  8. 8

    Zwartkruis, F. J., Wolthuis, R. M., Nabben, N. M., Franke, B. & Bos, J. L. EMBO J. 17, 5905–5912 (1998).

  9. 9

    York, R. D. et al. Nature 392, 622–626 (1998).

  10. 10

    Altschuler, D. L., Peterson, S. N., Ostrowski, M. C. & Lapetina, E. G. J. Biol. Chem. 270, 10373–10376 (1995).

  11. 11

    Franke, B., Akkerman, J. W. & Bos, J. L. EMBO J. 16, 252–259 (1997).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.