Role of Smad2/3 linker serine phosphorylation in regulating TGF-β signalling

Abstract

Transforming growth factor-β (TGF-β)/activin/nodal signalling play a critical role in many physiological and pathophysiological processes such as in embryogenesis, adult tissue homeostasis as well as in disease onset and progression. They signal via heteromeric complexes of type I and type II receptors, which phosphorylate and activate downstream signal effectors- Smad2 and Smad3 (Smad2/3). Phosphorylation of the C-terminal SxS motif of Smad2/3 by the receptors induces them to form oligomeric complexes with Smad4 and accumulate in the nucleus where they regulate transcription. In addition to the SxS phosphorylation, Smad2/3 can also be phosphorylated at their linker region. Smad2/3 are comprised of two highly conserved MH1 and MH2 domains connected by an unstructured linker that contains several phosphorylatable serine and threonine residues. Specifically, there are proline-directed linker threonine (LT) and serine (LS) residues that are phosphorylated by various kinases such as PI3K/mTORC2, MAPKs, Rho/ROCK, GSK3-β, CDKs etc. to regulate Smad2/3 activity by altering their subcellular localisation, transcription, and protein stability. LT and LS are closely linked and are perceived to have similar functions in TGF-β signalling. However, a study in our lab showed that they are differentially regulated, suggesting they may also have different functions in TGF-β signalling. Therefore, this study aims to dissect the individual roles of LS and LT in regulating TGF-β signalling by using various cell models such as human embryonic stem cells (hESCs), mouse embryonic stem cells (mESCs), and PC3 cells for mechanistic and functional studies. LS was found to be phosphorylated by CDK8/9 while LT was not much affected. Inhibition of LS phosphorylation (pLS) delayed the deactivation of Smad2, resulting in their nuclear accumulation and augmented transcriptional activity. Conversely, inhibition of pLS in inactivated Smad2 after withdrawing the agonist resulted in Smad2 proteasomal degradation. These findings reveal two distinct mechanisms by which pLS can regulate TGF-β signalling, which will greatly impact our understanding of the effects of non-canonical Smad signalling on canonical TGF-β signalling and also provides potential insights into ways TGF-β responses could be manipulated for therapeutic benefits.