Rac is required for growth cone function but not neurite assembly

Recent work has suggested that rac1 and other members of the rho family of small GTP-binding proteins play an important role in the formation of neural processes. We have explored the mechanism of this effect by comparing the spontaneous, growth cone-mediated growth and experimental tension-induced growth of axons in normal PC12 cells and in mutant cells expressing a dominant negative form of rac. PC12 that have been primed by exposure to NGF, but not naive PC12 cells, initiate a microtubule-rich process de novo in response to tension applied to cell body. As in chick sensory neurons, neurite elongation rate is proportional to applied tension above a threshold. Addition of cyclic AMP, which has been shown to rapidly augment NGF-induced neurite outgrowth in PC12, causes a rapid increase in the rate of neurite elongation at a given tension level. Expression of a dominant negative form of rac1 inhibits spontaneous, growth cone-mediated neurite elongation in response to NGF, but does not substantially affect tension-induced neurite elongation. That is, rac-deficient cells show a normal linear relationship between applied tension and elongation rate and the elongations contain a normal density of axial microtubules by immunofluorescent assay. Thus, rac1 is apparently required for the mechanisms that normally generate tension in an elongating neurite, but if this tension is provided from an outside source, then axonal elongation can proceed normally in rac1-deficient cells. We conclude that rac1 is required for the adhesive and motile function of growth cones rather than the assembly of neurites per se.