The Raf-l serine/threonine protein kinase requires phosphorylation of the s
erine at position 338 (S338) for activation. Ras is required to recruit Raf
-l to the plasma membrane, which is where S338 phosphorylation occurs. The
recent suggestion that Pak3 could stimulate Raf-l activity by directly phos
phorylating S338 through a Ras/phosphatidylinositol 3-kinase (P13-K)/-Cdc42
-dependent pathway has attracted much attention. Using a phospho-specific a
ntibody to S338, we have reexamined this model. Using LY294002 and wortmann
in, inhibitors of P13-K, we find that growth factor-mediated S338 phosphory
lation still occurs, even when P13-K activity is completely blacked. Althou
gh high concentrations of LY294002 and wortmannin did suppress S338 phospho
rylation, they also suppressed Ras activation. Additionally, we show that P
ak3 is not activated under conditions where S338 is phosphorylated, but whe
n Pak3 is strongly activated, by coexpression with V12Cdc42 or by mutations
that make it independent of Cdc42, it did stimulate S338 phosphorylation.
However, this occurred in the cytosol and did not stimulate Raf-l kinase ac
tivity. The inability of Pak3 to activate Raf-1 was not due to an inability
to stimulate phosphorylation of the tyrosine at position 341 but may be du
e to its inability to recruit Raf-l to the plasma membrane. Taken together,
our data show that growth factor-stimulated Raf-l activity is independent
of P13-K activity and argue against Pak3 being a physiological mediator of
S338 phosphorylation in growth factor-stimulated cells.