Rap1-mediated activation of extracellular signal-regulated kinases by cyclic AMP is dependent on the mode of Rap1 activation

Like other small G proteins of the Ras superfamily, Rap1 is activated by distinct guanine nucleotide exchange factors (GEFs) in response to different signals to elicit cellular responses. Activation of Rapt by cyclic AMP (cAMP) can occur via cAMP-dependent protein kinase A (PKA)-independent and PKA-dependent mechanisms. PKA-independent activation of Rapt1 by cAMP is mediated by direct binding of cAMP to Rap1-guanine nucleotide exchange factors (Rap1-GEFs) Epac1 (exchange protein directly activated by cAMP 1) and Epac2 (Epac1 and Epac2 are also called cAMP-GEFI and -GEFII). The availability of cAMP analogues that selectively activate Epacs, but not PKA, provides a specific tool to activate Rap1. It has been argued that the inability of these analogues to regulate extracellular signal-regulated kinases (ERKs) signaling despite activating Rap1 provides evidence that Rapt is incapable of regulating ERKs. We confirm that the PKA-independent activation of Rap1 by Epac1 activates a perinuclear pool of Rapt and that this does not result in ERK activation. However, we demonstrate that this inability to regulate ERKs is not a property of Rapt but is rather a property of Epacs themselves. The addition of a membrane-targeting motif to Epac1 (Epac-CAAX) relocalizes Epac1 from its normal perinuclear locale to the plasma membrane. In this new locale it is capable of activating ERKs in a Rap1- and cAMP-dependent manner. Rapt1 activation by Epac-CAAX, but not wild-type Epac, triggers its association with B-Raf. Therefore, we propose that its intracellular localization prevents Epac1 from activating ERKs. C3G (Crk SH3 domain Guanine nucleotide exchanger) is a Rapt exchanger that is targeted to the plasma membrane upon activation. We show that C3G can be localized to the plasma membrane by cAMP/PKA, as can Rapt when activated by cAMP/PKA. Using a small interfering RNA approach, we demonstrate that C3G is required for the activation of ERKs and Rap1 by cAMP/PKA. This activation requires the GTP-dependent association of Rap1 with B-Raf. These data demonstrate that B-Raf is a physiological target of Rap1, but its utilization as a Rap1 effector is GEF specific. We propose a model that specific GEFs activate distinct pools of Rap1 that are differentially coupled to downstream effectors.