Correction to Sequence Range
The authors have alerted the Science Signaling Editors that the sequence range of human PlexinB1 that was amplified by polymerase chain reaction was incorrectly stated in the Materials and Methods section. The correct sequence range is 1515-2135.
Ras and Rap GAP Function and GTPase Sequestration in Plexin-Mediated Cell Signaling Mechanisms
In a recent Science Signaling paper, Zhang and colleagues report the discovery of a Rap1 GAP function for plexins. This is an important paper. However, two aspects of the report require further study.
Ras GAP and Rap GAP function: Plexin transmembrane receptors function in diverse developmental and disease processes, including axon guidance, angiogenesis, and metastatic cell migration. The receptors are remarkable in that they interact with small GTPases at three levels: as effectors, binding to Rho GTPases; as regulators of Rho GTPase GAPs and GEFs; and as GAPs themselves. A GAP function towards Ras proteins was initially inferred from the homology of plexin intracellular regions to the Ras GAP family of regulatory proteins. GAP activity has been well established for members of all four human plexin families against R-Ras and M-Ras in cellular assays by five laboratories (1-3, for example), including by a biochemical assay using cell lysates (4). In these experiments, the Rho-family GTPase Rnd1 and clustering of plexin are required to activate plexin Ras GAP activity. Although R-Ras binding to the plexin-B1 intracellular region has been reported in vitro (5), it has been difficult to reconstitute the GAP activity for M- or R-Ras with purified protein components in solution (5, 6). Further work is clearly needed to successfully reconstitute the system.
Zhang and colleagues report a GAP function of plexin against Rap GTPases (6). In contrast to the situation with Ras, this Rap GAP activity is found in aqueous solution at high GTPase concentration. In particular, the intracellular domains of plexin-C1 and -D1 are functional, whereas plexin-B1 shows more modest activity. Activity could be stimulated in A-family plexins, and was increased in other plexins by induced N-terminal dimerization. Using constitutively active Rap1 protein, cell collapse was abolished in Sema3A-stimulated neurons, suggesting that the Rap GAP activity extends to cellular settings (6), but the signaling networks downstream of Rap in the case of plexins are not yet known [(7), for example]. Intriguingly, in vitro GAP activity seems to correlate inversely with the dimerization propensity of plexin family members themselves. For instance, plexin-A members can dimerize through their extracellular (8) as well as their intracellular domains (9, 10), whereas the Rho GTPase binding domains of plexin-C1 and -D1 are monomeric (10), suggesting that A- and B-family members can also engage in a dimer formation that inhibits Rap GAP activity. Overall the results suggest that plexins are multifunctional GAP proteins, similar to CAPRI, which is a dual specificity Ras/RAP enzyme (11). Similarly, we expect that different plexin GAP activities would also be controlled by different mechanisms and environments.
GTPase sequestration: The role and regulation of GTPase-plexin interactions also presents an ongoing area of intensive research with controversial findings and interpretations. Rho GTPases, especially Rnd1 and Rac1, can play an active role in receptor activation (1-4), although the structural mechanism is not yet clear, with experimental evidence supporting several, not necessarily exclusive models (3, 5, 6, 9). Structural perturbations in plexin crystallographic snap-shots that accompany GTPase binding may appear to be slight, but thermodynamic and mutagenesis studies suggest several binding modes and greater functional specificity for Rho GTPases than originally anticipated (12, 13, 14, 15). Intriguingly, Rap and Ras as well as several of the Rho GTPases also antagonize each other's function. Other modes of plexin-GTPase associations (referred to as sequestration) could also play a role. For example, as pointed out by Zhang and colleagues (6), there are specific cellular settings in which plexin binds R-Ras but, in absence of Rnd1, does not stimulate R-Ras GTP hydrolysis (16). This suggests a Ras sequestration model in which plexin may serve to withdraw the GTPase from the integrin system, thus diminishing cell adhesion in cell collapse. However, it is not clear whether sequestration provides an adequate control of Ras or Rho activity; in other systems compartmentalized or scaffold-bound proteins may still be active. Also, the binding affinity would need to be substantial and plexin concentration should be higher than that of the GTPases (12); neither of these requirements may be fulfilled in cells. Although sequestration is thought to have a larger impact on regulatory proteins or enzymes, recent studies also suggests that GTPase cycles are impacted when the GTPases or GTPase effectors are sequestered [for example, (17-20)]. The mechanisms are undoubtedly complex, but multilevel participation of plexins and GTPases is likely required in a temporally and spatially organized manner in order to create the waves of polymerization and depolymerization of the actin cytoskeleton, as well as integrin-mediated adhesion and de-adhesion, that are integral for cell migration. A "one fits-all" model, such as a general GTPase sequestration mechanism, is unlikely to work and much remains to be learned about GTPase-plexin interactions.
Response to Negishi and Buck
We are pleased that Drs. Manabu Negishi and Matthias Buck considered our work (1) important, and did not raise any specific criticisms of our data or interpretation.
As pointed out by Drs. Negishi and Buck, many small GTPases participate in plexin signaling. Some of them do so by direct interaction, others use indirect mechanisms. We agree that additional studies are required to reconcile some of the perplexing issues and elucidate these complicated signaling pathways, which should motivate future studies of people in the field, including ourselves.
Science Signaling. ISSN 1937-9145 (online), 1945-0877 (print). Pre-2008: Science's STKE. ISSN 1525-8882