Figure 2. Structure of Smad proteins. The R and Co-Smads show considerable homology within both the MH1 and MH2 domains. The MH domains are separated by a variable proline-rich linker region, which, in the case of Smad4, contains a Smad activation domain (SAD). **Smad2 is unable to bind directly to DNA, due to a 30-amino-acid insertion within exon3 of the MH1 domain. I-Smads have C-terminal domains structurally similar to R and Co-Smads. The N-terminus, however, fails to show significant homology.