TGF-beta receptor signaling is downregulated by proteasome and lysosome-mediated degradation of ubiquitinated TGFBR1, SMAD2 and SMAD3, as well as by dephosphorylation of TGFBR1, SMAD2 and SMAD3.

In the nucleus, SMAD2/3:SMAD4 complex stimulates transcription of SMAD7, an inhibitory SMAD (I-SMAD). SMAD7 binds phosphorylated TGFBR1 and competes with the binding of SMAD2 and SMAD3 (Hayashi et al. 1997, Nakao et al. 1997). Binding of SMAD7 to TGBR1 can be stabilized by STRAP, a protein that simultaneously binds SMAD7 and TGFBR1 (Datta et al. 2000). BAMBI simultaneously binds SMAD7 and activated TGFBR1, leading to downregulation of TGF-beta receptor complex signaling (Onichtchouk et al. 1999, Yan et al. 2009).

In addition to competing with SMAD2/3 binding to TGFBR1, SMAD7 recruits protein phosphatase PP1 to phosphorylated TGFBR1, by binding to the PP1 regulatory subunit PPP1R15A (GADD34). PP1 dephosphorylates TGFBR1, preventing the activation of SMAD2/3 and propagation of TGF-beta signal (Shi et al. 2004).

SMAD7 associates with several ubiquitin ligases, SMURF1 (Ebisawa et al. 2001, Suzuki et al. 2002, Tajima et al. 2003, Chong et al. 2010), SMURF2 (Kavsak et al. 2000, Ogunjimi et al. 2005), and NEDD4L (Kuratomi et al. 2005), and recruits them to phosphorylated TGFBR1 within TGFBR complex. SMURF1, SMURF2 and NEDD4L ubiquitinate TGFBR1 (and SMAD7), targeting TGFBR complex for proteasome and lysosome-dependent degradation (Ebisawa et al. 2001, Kavsak et al. 2000, Kuratomi et al. 2005). The ubiquitination of TGFBR1 can be reversed by deubiquitinating enzymes, UCHL5 (UCH37) and USP15, which may be recruited to ubiquitinated TGFBR1 by SMAD7 (Wicks et al. 2005, Eichhorn et al. 2012).

Basal levels of SMAD2 and SMAD3 are maintained by SMURF2 and STUB1 ubiquitin ligases. SMURF2 is able to bind and ubiquitinate SMAD2, leading to SMAD2 degradation (Zhang et al. 2001), but this has been questioned by a recent study of Smurf2 knockout mice (Tang et al. 2011). STUB1 (CHIP) binds and ubiquitinates SMAD3, leading to SMAD3 degradation (Li et al. 2004, Xin et al. 2005). PMEPA1 can bind and sequester unphosphorylated SMAD2 and SMAD3, preventing their activation in response to TGF-beta signaling. In addition, PMEPA1 can bind and sequester phosphorylated SMAD2 and SMAD3, preventing formation of SMAD2/3:SMAD4 heterotrimer complexes (Watanabe et al. 2010). A protein phosphatase MTMR4, residing in the membrane of early endosomes, can dephosphorylate activated SMAD2 and SMAD3, preventing formation of SMAD2/3:SMAD4 complexes (Yu et al. 2010).