Normal signaling by TGFbeta, in the absence of serum or exogenous factors, involves a rapid activation of Ras, Erks, and Sapks in proliferating cultures of TGFbeta-sensitive untransformed epithelial cells and human carcinoma cells. Expression of either RasN17 or dominant-negative (DN) MKK4, or addition of the MEK1 inhibitor PD98059, can block the ability of TGFbeta to induce AP-1 complex formation at the TGFbeta(1) promoter and to autoinduce its own production. The primary components present in this TGFbeta-stimulated AP-1 complex are JunD and Fra-2, although c-Jun, and possibly Fos B, may also be present. While there are two potential Smad binding elements (SBE's) in the TGFbeta(1) promoter, supershift assays suggest that at least one of these does not bind Smad4, and the other is unable to bind factors activated by TGFbeta. In contrast, TGFbeta autoinduction is Smad3-dependent, as DN Smad3 inhibits the ability of TGFbeta to stimulate TGFbeta(1) promoter activity. Our results indicate that TGFbeta can activate both the MKK4/Sapk and MEK/Erk pathways, through Ras and TGFbeta R(I) and R(II), to induce TGFbeta(1) production; Smad4 does not appear to be involved, and Smad3 appears to function independently of this Smad4. We also demonstrate that activation of the Ras/Mapk pathway by TGFbeta positively modulates Smad1-signaling-pathway activation by TGFbeta. In addition, Smad1 could enhance TGFbeta activation of the SBE reporter SBE-luc and this effect could be blocked by co-expression of a DN TGFbeta R(I) receptor or by the MEK1 inhibitor PD98059. This cross-talk between the MEK/Erk and Smad1 pathways was mediated through the four Erk consensus phosphorylation sites in the linker region of Smad1. Mutation of these sites resulted in a loss of the ligand-dependence of both Smad1-Smad4 interactions and nuclear accumulation of Smad1, as well as a loss of the ability of Smad1 to enhance TGFbeta-mediated SBE activation. Our results provide evidence that Erk-mediated phosphorylation of Smad1 in response to TGFbeta is critical for regulating Smad1 subcellular localization; this may be a key determinant in maintaining TGFbeta-dependent transcriptional activation.