The hallmarks of Rho-family-GTPase activation — stress fibres, lamellipodia and filopodia — require the Wiskott–Aldrich-syndrome protein (WASP) family and, of course, the actin-nucleating Arp2/3 complex. Cdc42 interacts directly with the WASP-family protein N-WASP to stimulate actin polymerization, but no such direct binding has been detected between Rac and the WASP-related protein WAVE. Indeed, unlike WASP, WAVE lacks a Cdc42/Rac-interactive binding (CRIB) motif, so how it functions downstream of Rac to induce membrane ruffles has been unclear. Now, however, research from Marc Kirschner's laboratory provides an explanation of how Rac and the adaptor protein Nck activate actin nucleation.

In contrast to N-WASP, which is autoinhibited by an intramolecular interaction between the carboxyl and amino termini, recombinant WAVE1 can constitutively induce actin nucleation by Arp2/3. To prevent actin nucleation occurring without external signals, WAVE activity in vivo must either be inhibited by other proteins or regulated post-translationally. To investigate the former possibility, Kirschner and colleagues looked for WAVE-interacting proteins, and used size-exclusion chromatography to analyse a protein complex from bovine brain extracts in which WAVE1 elutes. They identified four proteins in addition to WAVE1: a p53-inducible messenger RNA with a relative molecular mass of 140,000 (PIR121); a NCK-associated protein with a relative molecular mass of 125,000 (Nap125); a protein similar to hNap1-binding protein; and HSPC300.

The complex does not activate the Arp2/3 complex in vitro. Adding GTPγS-bound Rac1 (a non-hydrolysable form of GTP), however, does stimulate the activity of Arp2/3, suggesting that Rac1 can relieve WAVE1 inhibition in the complex. Both Nap125 and PIR121 can associate with NCK, so the authors tested whether NCK could induce the WAVE1 complex to activate Arp2/3, which it could. Furthermore, GTPγS–Rac1 and NCK activated the WAVE1 complex additively, but, because neither could induce additional activation of recombinant WAVE1, the authors proposed that they function to relieve inhibition rather than to stimulate activity.

This being the case, Rac1 and NCK might be able to induce dissociation of the complex to release active WAVE1. The complex indeed disassembled into two subcomplexes. Because the fraction that contained WAVE1 (WAVE1–HSPC300) stimulated Arp2/3-dependent actin nucleation, the scenario seems to be that Rac1 or NCK can induce the formation of lamellipodia/membrane ruffles by disassembling the trans-inhibited WAVE complex, leaving WAVE free to go about its duties.