Smads (or SMADs) comprise a family of structurally similar proteins that are the main signal transducers for receptors of the transforming growth factor beta (TGF-B) superfamily, which are critically important for regulating cell development and growth. The abbreviation refers to the homologies to the Caenorhabditis elegans SMA ("small" worm phenotype) and MAD family ("Mothers Against Decapentaplegic") of genes in Drosophila.

There are three distinct sub-types of Smads: receptor-regulated Smads (R-Smads), common partner Smads (Co-Smads), and inhibitory Smads (I-Smads). The eight members of the Smad family are divided among these three groups. Trimers of two receptor-regulated SMADs and one co-SMAD act as transcription factors that regulate the expression of certain genes.

The R-Smads consist of Smad1, Smad2, Smad3, Smad5 and Smad8/9, and are involved in direct signaling from the TGF-B receptor.

Smad4 is the only known human Co-Smad, and has the role of partnering with R-Smads to recruit co-regulators to the complex.

Finally, Smad6 and Smad7 are I-Smads that work to suppress the activity of R-Smads. While Smad7 is a general TGF-B signal inhibitor, Smad6 associates more specifically with BMP signaling. R/Co-Smads are primarily located in the cytoplasm, but accumulate in the nucleus following TGF-β signaling, where they can bind to DNA and regulate transcription. However, I-Smads are predominantly found in the nucleus, where they can act as direct transcriptional regulators.

Before Smads were discovered, it was unclear what downstream effectors were responsible for transducing TGF-B signals. Smads were first discovered in Drosophila, in which they are known as mothers against dpp (Mad), through a genetic screen for dominant enhancers of decapentaplegic (dpp), the Drosophila version of TGF-B. Studies found that Mad null mutants showed similar phenotypes to dpp mutants, suggesting that Mad played an important role in some aspect of the dpp signaling pathway.

A similar screen done in the Caenorhabditis elegans protein SMA (from gene sma for small body size) revealed three genes, Sma-2, Sma-3, and Sma-4, that had similar mutant phenotypes to those of the TGF-B like receptor Daf-4. The human homologue of Mad and Sma was named Smad1, a portmanteau of the previously discovered genes. When injected into Xenopus embryo animal caps, Smad1 was found to be able to reproduce the mesoderm ventralizing effects that BMP4, a member of the TGF-B family, has on embryos. Furthermore, it was demonstrated that Smad1 had transactivational ability localized at the carboxy terminus, which can be enhanced by adding BMP4. This evidence suggests that Smad1 is responsible in part for transducing TGF-B signals.

Smads are roughly between 400 and 500 amino acids long, and consist of two globular regions at the amino and carboxy termini, connected by a linker region. These globular regions are highly conserved in R-Smads and Co-Smads, and are called Mad homology 1 (MH1) at the N-terminus, and MH2 at the C-terminus. The MH2 domain is also conserved in I-Smads. The MH1 domain is primarily involved in DNA binding, while the MH2 is responsible for the interaction with other Smads and also for the recognition of transcriptional co-activators and co-repressors. R-Smads and Smad4 interact with several DNA motifs though the MH1 domain. These motifs include the CAGAC and its CAGCC variant, as well as the 5-bp consensus sequence GGC(GC)|(CG). Receptor-phosphorylated R-Smads can form homotrimers, as well as heterotrimers with Smad4 in vitro, via interactions between the MH2 domains. Trimers of one Smad4 molecule and two receptor-phosphorylated R-Smad molecules are thought to be the predominant effectors of TGF-β transcriptional regulation. The linker region between MH1 and MH2 is not just a connector, but also plays a role in protein function and regulation. Specifically, R-Smads are phosphorylated in the nucleus at the linker domain by CDK8 and 9, and these phosphorylations modulate the interaction of Smad proteins with transcriptional activators and repressors. Furthermore, after this phosphorylation step, the linker undergoes a second round of phosphorylations by GSK3, labelling Smads for their recognition by ubiquitin ligases, and targeting them for proteasome-mediated degradation. The transcription activators and the ubiquitin ligases both contain pairs of WW domains. These domains interact with the PY motif present in the R-Smad linker, as well as with the phosphorylated residues located in the proximity of the motif. Indeed, the different phosphorylation patterns generated by CDK8/9 and GSK3 define the specific interactions with either transcription activators or with ubiquitin ligases. Remarkably, the linker region has the highest concentration of amino acid differences among metazoans, although the phosphorylation sites and the PY motif are highly conserved.