The Ben May Institute for Cancer Research
The University of Chicago
Chicago, IL, USA
Samuel Lunenfeld Research Institute
Mt. Sinai Hospital
Toronto, Ontario, Canada
Intracellular signaling networks control many different aspects of cellular behavior, and are assembled through the interactions of proteins with one another, and with lipids, nucleic acids and small molecules. This design allows the internal ‘wiring’ of the cell to respond dynamically to extracellular signals, and provides flexibility so that a limited set of signaling proteins can provide distinct biological functions in different cellular contexts. For example, mammalian receptors that are important in axon guidance, and formation of the mammalian brain, can also regulate post-synaptic functions associated with learning and memory after birth. Investigating how such signaling pathways are organized is relevant not only for understanding how normal cells work, but also for appreciating the molecular basis for disease, since many human disorders result from breakdowns in signal transduction.
Post-translational protein modifications, most notably phosphorylation by protein kinases, represent a common mechanism through which signaling systems are controlled. This regulatory device is inseparably linked with the physical interactions of signaling proteins, since protein phosphorylation commonly exerts its effects by creating binding sites for a set of protein interaction domains, whose ability to bind their targets is phosphorylation-dependent.
Protein-protein interactions recruit cytoplasmic polypeptides to activated receptors, direct their assembly into larger complexes, target them to defined subcellular locations, and determine the specificity with which enzymes interact with their substrates. Typically, 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 often closely juxtaposed in space, while the binding surface lies on the opposite face of the domain. This arrangement gives rise to a ‘plug-and-play’ architecture that allows the domain to be inserted into a host polypeptide such that its ligand-binding site projects to engage ligand and maintain binding function. In evolutionary terms, this presents an attractive model by which polypeptides gain new functions and increased complexity, through the accretion of modular domains or exposed ligands.
Phospho-dependent protein interaction domains typically recognize specific peptide motifs on their binding partners, in a fashion that depends on the phosphorylation of a tyrosine or serine/threonine residue in the recognition sequence. The mechanisms by which protein kinases and phospho-dependent interaction domains synergize to activate biochemical pathways is illustrated by receptor tyrosine kinase (RTK) signaling. RTKs are often activated by growth factor-mediated dimerization (or by oncogenic mutations which induce constitutive oligomerization), which result in the cross-phosphorylation of one receptor chain by its neighbour. This autophosphorylation has two consequences. Phosphorylation within the activation segment of the kinase domain results in a conformational change that stimulates catalytic activity, while phosphorylation of tyrosine residues in non-catalytic regions (e.g. the juxtamembrane region, C-terminal tail or kinase insert) or the activation segment of the kinase domain create binding sites for the SH2 domains or PTB domains of cytoplasmic targets. The human genome is currently estimated to encode 120 SH2 domains, which are found in 110 distinct proteins. SH2 domains bind their ligands as an extended strand; they all recognize phosphotyrosine, through a conserved, basic binding pocket, and also bind at least three residues immediately C-terminal to the phosphotyrosine, in a fashion that differs from one SH2 domain to another and provides an element of specificity in signal transduction. Thus the sequence contexts of a receptor’s autophosphorylation sites determine the identities of the SH2-containing proteins that bind the activated RTK, and the spectrum of signaling pathways activated in the cell. Consistent with the view that the SH2 domain serves as a portable module to couple phosphotyrosine signals to intracellular targets, SH2-containing proteins can have a wide range of biological functions, including the regulation of Ras-like GTPases, phospholipid metabolism, gene expression, cytoskeletal organization, and protein phosphorylation. In many cases, SH2 domains are not covalently linked to catalytic modules, but rather are found in adaptor proteins, composed exclusively of interaction sequences, such as SH2 and SH3 domains. Such adaptors can nucleate the formation of multi-protein signaling complexes that direct phosphotyrosine signals to particular intracellular targets.
Phosphotyrosine-containing motifs are also recognized by PTB domains, found in docking proteins such as IRS-1, a principal substrate of the insulin-receptor. PTB domains typically bind NPXY sequences, which form B-turns; PTB domains of proteins involved in tyrosine kinase signaling (e.g. IRS, Shc, FRS2 and Dok family members) require phosphorylation of the NPXY tyrosine for stable binding. Such docking proteins themselves possess multiple tyrosine phosphorylation sites, which engage SH2-containing proteins, exemplifying how a succession of phospho-dependent protein-protein interactions can be used to construct signaling pathways and networks. Although SH2 and PTB domains both recognize phosphotyrosine containing sequences, they are structurally unrelated, and engage their phospho-peptide ligands in different ways (albeit that basic arginine and lysine residues are a common feature of their phosphotyrosine-binding pockets). Similarly, a growing number of interaction domains share an ability to selectively bind phosphoserine/threonine-containing motifs, but have quite different structural folds.
The majority of protein kinases in eukaryotes phosphorylate serine/threonine residues, and regulate facets of cellular function ranging from the cell cycle to gene expression and metabolism. 14-3-3 proteins, FHA domains, MH2 domains, WD40 repeat domains, BRCT domains, Polo-Box domains, FF domains and WW domains all have the ability to bind specific phosphoserine/ threonine-containing peptide motifs. FHA domains, for example, are found in protein kinases involved in DNA damage repair, and recognize phosphothreonine, with selectivity being provided by recognition of the +3 residue. The interactions mediated by the FHA domains of the Rad53 protein kinase in yeast are required for the cellular response to DNA damage. MH2 domains are structurally similar to FHA domains, and are found in SMAD proteins, that serve as targets for activated TGFβ receptor serine/threonine kinases (RSK). Autophosphorylation of the serine-rich juxtamembrane region of the type I TGFβ receptor creates a binding site for the SMAD MH2 domain, which recognizes pSer-X-pSer motifs; the receptor-bound SMAD is subsequently phosphorylated within a C-terminal motif, leading to the formation of a phospho-dependent SMAD complex that leaves the receptor and moves to the nucleus to regulate gene expression. Although the details are quite different, there are significant parallels in the recognition of specific phosphorylated motifs of activated RTKs and RSKs by interaction domains on their targets.
Recognition of protein phosphorylation can lead to a complex interplay with other cellular systems. For example, protein ubiquitylation, and resulting proteolysis or receptor internalization, is frequently associated with phosphorylation of the target protein. Tyrosine phosphorylated receptors are recognized by the variant SH2 domain of the E3 protein ubiquitin ligase c-Cbl, which recruits an E2 ubiquitin ligase and induces receptor monoubiquitylation. The modified receptor is in turn recognized by proteins with ubiqutuin-binding domains involved in endocytosis. Similarly, serine/threonine phosphorylation of proteins destined for degradation can create binding sites for the WD40 repeat or leucine rich repeat domains of F-box proteins, the targeting subunits of SCF E3 ubiquitin ligase complexes. Typically this results in polyubiquitination and recruitment to the proteosome.
As a final word, it is interesting to note that although protein interaction domains are superficially rather simple in their binding properties, increasing evidence suggests that they have been selected for their flexibility, their ability to assemble multiprotein machines, and their potential to mediate sophisticated biological functions such as setting thresholds, integrating signals and creating all-or-none switches. Phosphorylation may be viewed as a device to control the assembly of protein complexes, and thereby to regulate the dynamic behavior of the cell. Phospho-dependent interaction domains, taken together with other interaction domains, provide a highly adaptable and dynamic platform for cellular regulation.
CST gratefully acknowledges the following contributors to the Protein Domains section of our catalog and website:
Piers Nash1, Dan Lin3, Kathleen Binns2, Clark Wells2, Rob Ingham2, Terry Kubiseski2, Bernard Liu1, Matt Smith2,3, Ivan Blasutig2,3, Maria Sierra1, Caesar Lim2,3, Michael Arcé1, Jim Fawcett2 and Tony Pawson2,3.