A number of interaction domains recognize short peptide motifs in the surface loops or intrinsically unstructured regions of proteins, in a fashion that is regulated by post-translational modification of the peptide ligand. Regulation of protein–protein interactions in this way is now recognized as a rather general mechanism for regulating cellular processes, but it was originally identified in the context of Src homology 2 (SH2) domains which we briefly discussed in Chapter 3 (see section 3-1) and which are contained in many cell signaling proteins. Src is the name of a protein-tyrosine kinase first discovered as the product of an oncogene that causes sarcomas in chickens (hence Src). The SH2 domain is one of its two regulatory domains and recognizes peptide sequences that contain a phosphotyrosine. In this section we describe the recognition of phosphotyrosine motifs by SH2 domains, which is particularly well studied; and in the next sections we describe other interaction domains that function in signaling through selective binding to modified sites.

SH2 domains interact with specific phosphotyrosine-containing peptides through their ability to recognize both the phoshotyrosine residue and the side chains of the adjacent amino acids (usually located carboxy-terminal to the phosphotyrosine). SH2 domains are approximately 100 amino acids in length, and fold into a compact structure with a central beta sheet that separates the domain into two binding pockets (Figure 8-1.1), one of which is highly conserved among all SH2 domains and serves primarily to bind the phosphotyrosine while the other, the specificity pocket, is more variable and binds the side chains of the adjacent amino acids. Binding to the phosphotyrosine is through a bidentate ionic interaction with an essential argi-nine residue (the fifth residue in β strand B, and hence termed ArgβB5), and through H-bond, hydrophobic and amino-aromatic interactions with other amino acids. ArgβB5 sits at the base of the pocket, and is precisely positioned to interact with the phosphate group of a phosphotyrosine; the shorter side chains of phosphoserine or phosphothreonine would be unable to project far enough into the pocket to encounter this arginine, explaining the selectivity of SH2 domains for phosphotyrosine (Figure 8-1.2). Phosphopeptides typically bind an SH2 domain in an extended conformation, so that they run across the central beta sheet (see Figure 3-1.2), thereby positioning the carboxy-terminal amino acids to interact with the second, more variable binding pocket. In some cases SH2 domains engage only three carboxy-terminal residues (in the +1 to +3 positions), while others bind more extended phosphorylated motifs, up to eight amino acids in length, and can engage residues both amino- and carboxy-terminal to the phosphotyrosine. Different SH2 domains bind preferentially to distinct amino acids in the +1 to +3 positions of the phosphopeptide; for example the SH2 domain of the adaptor protein Grb2 selects phosphotyrosine sites with an asparagine at the +2 position, whereas the SH2 domains of the p85 subunit of phosphatidylinositol 3'-kinase (PI3K) prefer a methionine at the +3 position. The affinity with which a particular peptide binds a given SH2 domain will therefore depend on its phosphorylation, since about half of the binding energy comes from recognition of the phosphotyrosine, and also on the fit of the carboxy-terminal peptide residues with the specificity pocket of the SH2 domain. As a consequence of this structural arrangement, the sequence of amino acids flanking the phosphotyrosine sites of signaling proteins, such as activated growth factor receptors, influences which SH2-containing proteins are recruited to each phosphorylated motif.

Individual SH2 domains typically bind their preferred phosphorylated peptide motif with a K_d of ~1 micromolar, although in some cases the interactions can be tighter (~100 nM). They also show rapid on- and off-rates, indicating that SH2 domain–phosphopeptide interactions are highly dynamic, allowing signaling to be rapidly started and stopped.

As well as having distinct binding specificities, different SH2 domains are found in proteins containing other kinds of functional domains, which may be other types of interaction domains, as in adaptor proteins, or catalytic domains, as with some signaling kinases; or, often, both. They thus regulate a variety of processes including the GTP-binding of small GTPases, phospholipid metabolism, the assembly of the actin cytoskeleton, ubiquitination, tyrosine phosphorylation and transcription. Figure 8-1.3 illustrates a selection of the proteins that contain SH2 domains, with their functions. The cellular response to a tyrosine kinase signal is therefore determined by the ability of the resulting phosphotyrosine motifs to associate with specific SH2 domain proteins, and the downstream pathways these proteins activate or repress.

This principle can be seen in signaling from the insulin and epidermal growth factor (EGF) receptors. The autophosphorylated insulin receptor binds a scaffold protein, insulin receptor substrate (IRS)-1, which is phosphorylated by the receptor at several tyrosine residues that lie in YXXM motifs, and consequently bind the SH2 domains of the p85 subunit of PI3K, leading through the production of PI(3,4,5)P₃ to changes in cell growth and survival (Figure 8-1.4a; we discuss this pathway in more detail in Chapter 16). By contrast, the activated EGF receptor bears multiple pYXN motifs, both within its own autophosphorylated carboxy-terminal tail and in the associated scaffold protein Shc, and these bind the SH2 domain of Grb2 leading to activation of the Ras GTPase (see section 5-2) and its downstream targets such as the Erk MAP kinase pathway (Figure 8-1.4b).

Definitions

insulin receptor substrate-1 (IRS-1): a scaffold that mediates signaling from the insulin and IGF-1 receptors. IRS-1 has amino-terminal PH domain and PTB domains, that mediate its binding to the autophosphorylated receptor, and a carboxy-terminal region containing multiple YXXM sites that bind SH2 domains upon tyrosine phosphorylation.

IRS-1: see insulin receptor substrate-1.

phosphatidylinositol (PI) 3'-kinase (PI3K): a signaling enzyme that catalyzes the conversion of PI(4,5)P₂ to PI(3,4,5)P₃ by phosphorylation of the D-3 position of the inositol head group. PI3K has two subunits.

PI3K: see phosphatidylinositol 3'-kinase.

Shc: a scaffold protein with an amino-terminal PTB domain and a carboxy-terminal SH2 domain.

Src homology 2 (SH2) domain: a conserved sequence of approximately 100 amino acids that folds into a globular domain that binds phosphotyrosine-containing peptide sequences when expressed in isolation from its host protein.