Rosen was intrigued initially by the location of a tyrosine in WASP that others had found was phosphorylated. In structural models, Tyr 291 is buried in the fold that forms in the autoinhibisted form of WASP. Sure enough, Tyr 291 could only be phosphorylated (thus further activating WASP activity) when the structure was opened up via addition of the activating Cdc42.
Both Tyr 291 phosphorylation and Cdc42 activation can be driven by Src family kinases. But Cdc42 activation and opening of WASP take some time, so only a persistent Src signal will still be around to phosphorylate Tyr 291 by the time it is exposed.
WASP could then act as a memory device if Src maintains the Tyr 291 phosphorylation until after Cdc42 is turned off (as might happen if a GTPase activating protein drops by and shuts off Cdc42). Torres and Rosen showed that this form of WASP clasps and protects the phosphorylation on Tyr 291 so it is resistant to removal by phosphatases. In contrast to the initial, unphosphorylated form of WASP that took such a sustained signal to activate, the phosphorylated WASP is now in a primed form. A future signal need only tickle the Cdc42 system to achieve full activation.
For now, the site of action of these mechanisms are unknown. “What we've found is exclusively biochemical,” says Rosen. “We don't know where this will occur in vivo.” But, he says, “having this as the biochemical concept will drive the kind of work” needed to put the findings in context. Rosen is looking for places where later responses might be affected by earlier experience. The best candidate is in neurite outgrowth, where N-WASP phosphorylation has been seen to persist long after the kinase stimulus has vanished. Perhaps such phenomena allow neurites to keep extending even through areas that have temporary drops in outgrowth factors. ▪