Inhibition of Smad transcriptional activity and antiproliferative function by CDK phosphorylation

Abstract

2604

We demonstrate that Smad3, a key mediator of TGF-beta antiproliferative responses, is a major physiological substrate of G1 cyclin-dependent kinases CDK4 and CDK2. Except for the retinoblastoma protein (Rb) family, Smad3 is the only substrate demonstrated so far for CDK4, and Smad3 can be phosphorylated by CDK4 to a greater extent than Rb. We have mapped both the in vivo and the in vitro CDK4 and CDK2 phosphorylation sites in Smad3. These sites are phosphorylated by CDK4 and CDK2 in vitro and in a cell-cycle dependent manner in vivo. Moreover, we show that their phosphorylation is reduced in CDK4 knockout mouse embryonic fibroblasts (MEF), and their phosphorylation is increased in CDK4 R24C/R24C MEF, which have elevated CDK4 activity. In addition, RNAi experiments also indicate that CDK4 and CDK2 phosphorylate these sites. CDK phosphorylation of Smad3 inhibits its basal as well as TGF-beta induced transcriptional activity, thus preventing activation of the expression of CDK inhibitors p15 and p21. In addition, CDK phosphorylation of Smad3 decreases its ability to downregulate the expression of the protooncogene c-myc. Using Smad3-/- primary mouse embryonic fibroblasts and other epithelial cell lines, we further show that Smad3 inhibits cell cycle progression from the G1 to S phase and that CDK phosphorylation of Smad3 facilitates cell proliferation. Since cancer cells often contain high levels of CDK activities, inactivation of Smad3, and presumably the homologous Smad2 as well, by CDK phosphorylation likely contributes to tumorigenesis and TGF-β resistance in cancers.