An ever-increasing body of data suggests that proteins involved in the regulation of cellular events such as signal transduction, the cell cycle, protein trafficking, targeted proteolysis, cytoskeletal organization and gene expression are constructed in a modular fashion from a combination of interaction and catalytic domains. Interaction domains drive signaling polypeptides into specific multi-protein complexes, and thereby couple cell surface receptors to intracellular biochemical pathways that control cellular responses to external signals. The pathways and networks that link receptors to their ultimate targets frequently involve a series of protein-protein interactions, that recruit and confine signaling proteins to an appropriate subcellular location, and determine the specificity with which enzymes interact with their targets, as with association of protein kinases and their substrates. Typically, protein-protein interaction domains are independently folding modules of 35-150 amino acids, that can be expressed in isolation from their host proteins while retaining their intrinsic ability to bind their physiological partners. Their N- and C-termini are usually close together in space, while their ligand-binding surface lies on the opposite face of the domain. This arrangement allows the domain to be inserted into a host protein while projecting its ligand-binding site to engage another polypeptide.

Protein-protein interaction domains can be divided into separate families, that are related either by sequence or ligand-binding properties. Thus, a large number of cytoplasmic proteins contain one or two SH2 domains that directly recognize phosphotyrosine-containing motifs, such as those found on activated receptors for growth factors, cytokines and antigens. SH2 domains commonly recognize phosphotyrosine, but differ in their preference for the amino acids immediately following the phosphorylated residue, which provides an element of specificity in signaling by tyrosine kinases. Interaction domains are often used repeatedly in numerous different proteins to mediate a particular type of molecular recognition, and indeed the human genome is predicted to encode at least 120 SH2 domains. However, phosphotyrosine-containing motifs are also recognized by a quite different class of interaction modules, termed PTB domains, found on docking proteins such as the IRS-1 substrate of the insulin receptor. Furthermore, a growing family of interactions domains, including 14-3-3 proteins, FHA domains and WD40-repeat domains recognize specific phosphoserine/threonine motifs, and thereby mediate the biological activities of protein-serine/threonine kinases. Recent data suggest that other forms of post-translational protein modification control modular protein-protein interactions. Notably, acetylation or methylation of lysine residues on histones creates binding sites for the Bromo and Chromo domains, respectively, of proteins involved in chromatin remodeling. Taken together, these findings suggest that the dynamic control of cellular behavior exerted by covalent protein modifications is mediated by interaction domains, that regulate the associations of signaling proteins one with another.

A large group of interaction domains (SH3, WW, EVH1) bind proline-rich motifs; these complexes are less dependent on post-translational modifications, and therefore more stable than the phospho-dependent interactions involving SH2 domains. Similarly, PDZ domains bind the extreme C-termini of other polypeptides, such as ion channels and receptors, in a fashion that appears important for the localization of their targets to particular subcellular sites, as well as for downstream signaling.

The interactions discussed above all involve the ability of a folded interaction domain to recognize a short peptide motif. In addition, a number of modules form homo- or heterotypic domain-domain interactions. These include PDZ domains, which are rather versatile since they can form heterodimers as well as binding short C-terminal peptide motifs, and SAM domains.

In addition to interaction domains that engage specific peptide motifs, a growing number of modules have been identified that recognize selected phospholipids, notably phosphoinositides (PI). Strikingly, PH domains can bind either PI-4,5-P2 or PI-3,4,5-P3, and thereby mediate the effects of lipid kinases and phosphatases on cellular function. Such phospholipid-binding domains serve both to concentrate signaling proteins at specific subregions of the plasma membrane, and to regulate the enzymatic activities of their host proteins, either directly or by co-recruitment of another regulatory protein. Modules such as FYVE domains can recognize PI-3-P, and may play an important role in the trafficking of proteins within the cell.

Two features of interaction domains bear particular comment. One is their apparent versatility. Thus, although PTB domains were originally discovered through their ability to bind phosphotyrosine in the context of an Asn-Pro-X-Tyr (NPXY) motif which forms a b-turn, it appears that many PTB domains recognize NPXY-related peptide motifs, but in a phospho-independent fashion. Thus PTB domains likely evolved to bind unphosphorylated peptides, and have subsequently developed a capacity to recognize phosphotyrosine in a few specific cases. Furthermore, an individual PTB domain, such as those from the Numb and FRS-2 proteins, can recognize two quite different peptide ligands. Surprisingly, although PTB domains primarily bind peptide motifs and PH domains recognize phosphoinositdes, they have a very similar structural fold, which is shared by other interaction domains, including EVH1 domains which bind specific proline-rich sequences. It seems that the PH/PTB/EVH1 domain fold provides a framework that can be used for multiple distinct types of intermolecular interactions. Second, different interaction domains are frequently covalently linked within the same polypeptide chain, to yield a protein that can mediate multiple protein-protein and protein-phospholipid interactions. This modular organization of signaling proteins can then target proteins to the appropriate site within the cell, and direct their interactions with cell surface receptors and downstream targets. The reiterated and combinatorial use of interaction domains can in principle provide a wiring plan that controls and integrates the flow of information within the cell.

domains listed by category ... or domains at a glance...