- SH2, SH3 and PDZ domains are among the best-characterized protein–protein interaction domains in molecular and cellular biology. These domains allow proteins to recognize specific molecular features on their binding partners and assemble into organized signaling complexes. Although all three participate in protein–protein interactions, they recognize different structural or sequence features and therefore perform different molecular functions. Understanding SH2, SH3 and PDZ domains provides an important foundation for understanding how cells convert extracellular signals into intracellular responses and how proteins are organized into functional molecular networks.
- Protein–protein interaction domains are specialized regions of proteins that contribute to the recognition and binding of other proteins. Their binding specificity arises from the three-dimensional structure and chemical properties of the interaction surface. Some domains recognize short peptide sequences, whereas others recognize post-translational modifications or larger structural surfaces. SH2, SH3 and PDZ domains illustrate these different principles particularly well. SH2 domains commonly recognize phosphorylated tyrosine-containing sequences, SH3 domains commonly recognize proline-rich sequences, and PDZ domains frequently recognize short sequences at the C-termini of partner proteins.
- The SH2 domain, or Src homology 2 domain, is a compact protein domain best known for recognizing phosphorylated tyrosine residues. The discovery and characterization of SH2 domains helped establish an important principle in cell signaling: phosphorylation can function not only as a biochemical modification but also as a molecular docking signal. When a tyrosine residue on a signaling protein becomes phosphorylated, an SH2-containing protein can recognize the modified site and become recruited to that signaling complex.
- This mechanism is particularly important in pathways controlled by protein tyrosine kinases. Receptor tyrosine kinases and other tyrosine kinases can phosphorylate specific tyrosine residues on receptors or intracellular signaling proteins. These phosphorylated residues can then serve as binding sites for proteins containing SH2 domains. The recruitment of these proteins brings additional enzymatic or regulatory activities into the signaling complex and allows the original phosphorylation event to propagate through the cell.
- SH2-domain recognition is not determined simply by the presence of a phosphorylated tyrosine. The amino acids surrounding the phosphotyrosine can contribute strongly to binding specificity. Different SH2 domains therefore recognize different sequence contexts surrounding phosphorylated tyrosines. This allows a relatively small chemical modification—addition of a phosphate group—to generate highly specific molecular recognition signals.
- This principle creates a phosphorylation-dependent molecular switch. Before phosphorylation, a protein may not efficiently bind an SH2-containing partner. After phosphorylation, the modified residue can become a docking site. Removal of the phosphate group can then weaken or eliminate the interaction. Protein kinases and protein phosphatases therefore regulate not only the biochemical activity of proteins but also the composition of protein–protein interaction networks.
- SH2 domains are found in numerous signaling proteins. Some SH2-containing proteins are enzymes, whereas others function primarily as adaptor or scaffold proteins. Enzymatic SH2-containing proteins can be recruited to activated receptors and subsequently modify downstream targets. Adaptor proteins can use SH2 domains to recognize phosphorylated proteins while using additional interaction domains to recruit other components. In this way, one protein can function as a molecular bridge between different stages of a signaling pathway.
- An important example is the organization of signaling downstream of activated receptor tyrosine kinases. Following ligand binding, receptor activation can lead to receptor autophosphorylation. The phosphorylated receptor then provides docking sites for proteins containing appropriate interaction domains. SH2-containing proteins can bind these phosphotyrosine sites and recruit additional signaling components. The resulting multiprotein assembly can activate pathways controlling cell proliferation, differentiation, metabolism, survival and gene expression.
- The SH3 domain, or Src homology 3 domain, represents a different mechanism of protein recognition. SH3 domains commonly recognize proline-rich sequences, particularly sequences with characteristic proline-rich structural conformations. Rather than primarily recognizing a phosphorylated amino acid, SH3 domains generally recognize the sequence and structural properties of a peptide region in the interacting protein.
- SH3-mediated interactions are important in many cellular processes, including signal transduction, cytoskeletal organization, membrane trafficking and assembly of multiprotein complexes. Proteins containing SH3 domains can bind proline-rich regions in signaling proteins and thereby bring different components into physical proximity. This makes SH3 domains particularly useful for assembling dynamic molecular complexes.
- The proline-rich regions recognized by SH3 domains are often located within intrinsically disordered regions. Because these regions do not need to maintain one rigid three-dimensional structure before binding, they can contain multiple short interaction motifs. A protein may therefore use several disordered regions to interact with different partners or use multiple motifs to regulate the strength and timing of protein association.
- SH3 domains frequently occur together with other interaction domains. A signaling protein may contain an SH2 domain that recognizes a phosphorylated tyrosine and an SH3 domain that recognizes a proline-rich region of another protein. Such multidomain proteins can therefore connect phosphorylation-dependent recognition with additional protein–protein interactions. This modular architecture is one reason why signaling networks can rapidly assemble complex molecular structures from relatively small numbers of domain types.
- The combination of interaction domains is particularly important in adaptor proteins. An adaptor may lack a major catalytic activity but contain several domains that recognize different molecular partners. By binding one protein through one domain and another protein through a second domain, the adaptor brings the proteins together. This spatial organization can greatly influence the efficiency and direction of signaling.
- The PDZ domain provides another major example of a protein–protein interaction domain. PDZ domains are generally compact interaction modules that frequently recognize short peptide sequences located at the C-terminal ends of partner proteins. The name PDZ derives historically from three proteins in which related domains were identified: postsynaptic density protein 95, the Drosophila tumor suppressor protein Discs large and the tight junction protein ZO-1.
- PDZ domains are particularly important in organizing proteins at cellular membranes. Many membrane proteins contain short cytoplasmic C-terminal sequences that can be recognized by PDZ domains in scaffold proteins. The scaffold can therefore recruit receptors, ion channels, signaling enzymes and cytoskeletal components into organized molecular complexes.
- The ability of PDZ domains to recognize C-terminal sequences provides an effective mechanism for spatial organization. A membrane receptor may contain a particular C-terminal sequence that is recognized by a PDZ-containing scaffold. Binding of the receptor to the scaffold can determine its localization, stability, signaling properties or association with other proteins. PDZ-mediated interactions therefore contribute to the organization of signaling complexes rather than simply producing isolated protein–protein contacts.
- PDZ domains are especially prominent in neuronal signaling. Neurons contain highly specialized membrane regions where receptors, ion channels, signaling proteins and cytoskeletal components must be precisely organized. PDZ-containing scaffold proteins can help assemble these components into functional complexes at synapses and other specialized membrane domains. Similar principles operate in epithelial cells, where PDZ proteins contribute to the organization of proteins at cell junctions and polarized membrane surfaces.
- Although SH2, SH3 and PDZ domains have different recognition mechanisms, they share an important conceptual feature: each provides a modular molecular recognition system. The domain acts as a structural unit capable of recognizing particular features of another protein. Because these domains can be combined with catalytic domains, additional interaction domains or regulatory regions, proteins can perform multiple functions within the same signaling pathway.
- The specificity of these interactions depends on more than simply matching a domain with a generic peptide sequence. The amino acid sequence surrounding the recognized site, the three-dimensional conformation of the binding partners, electrostatic interactions, hydrophobic contacts and the local cellular environment can all influence binding. As a result, the same class of interaction domain can recognize many related partners while still displaying substantial differences in binding preference.
- Post-translational modifications are particularly important for regulating interaction domains. SH2 domains provide the clearest example because phosphorylation of tyrosine creates their principal recognition signal. However, post-translational modifications can also influence SH3- and PDZ-mediated interactions indirectly by changing the conformation, localization or accessibility of binding sites. Phosphorylation or other modifications can therefore alter protein interaction networks without changing the underlying amino acid sequence.
- The cellular concentration and localization of interacting proteins also influence these interactions. A protein containing an SH2, SH3 or PDZ domain may have many potential partners in the cell, but only some of them will be present at the same location and time. Consequently, protein–protein interactions are determined not only by biochemical compatibility but also by cellular organization.
- This principle is particularly important for transient signaling complexes. Many signaling interactions occur for only a limited period following receptor activation or another stimulus. A phosphorylated tyrosine can recruit an SH2-containing protein, for example, but dephosphorylation can terminate the interaction. Similarly, an interaction involving an SH3 or PDZ domain can be regulated by changes in protein concentration, localization, modification or competition with other binding partners.
- The distinction between binding affinity and biological specificity is important here. A domain may bind a peptide strongly in an isolated biochemical experiment, but the interaction may not occur extensively inside a cell because the proteins are separated into different compartments or because competing interactions are present. Conversely, several individually modest interactions can cooperate to produce a strong and biologically important association within a multiprotein complex.
- This cooperative behavior is related to multivalency. A protein may contain several interaction domains, while another protein contains several corresponding binding motifs. The resulting multipoint interaction can produce a stable complex even when each individual interaction is relatively weak. Multivalent interactions are common in signaling and scaffold proteins and contribute to the formation of dynamic molecular assemblies.
- SH2, SH3 and PDZ domains also illustrate the relationship between protein domains and short linear motifs. The domain is usually a relatively compact structural unit, whereas the binding sequence recognized by the domain can be much smaller. For example, an SH2 domain recognizes a phosphotyrosine-containing sequence, an SH3 domain recognizes a proline-rich motif and a PDZ domain often recognizes a short C-terminal sequence. The interaction therefore involves cooperation between a folded recognition domain and a smaller sequence feature in the partner protein.
- This domain–motif relationship is particularly important for proteins containing intrinsically disordered regions. A disordered region can contain several short linear motifs that serve as docking sites for different interaction domains. Because the region is flexible, the same protein can potentially interact with multiple partners in alternative complexes. Such proteins are therefore well suited to participate in dynamic signaling networks.
- The modular nature of these interaction systems also contributes to protein evolution. New protein functions can emerge through duplication, recombination or modification of existing interaction domains and motifs. A protein can acquire a new interaction domain and consequently gain the ability to associate with a new signaling partner. Likewise, changes in a short interaction motif can alter partner specificity without requiring major changes throughout the entire protein.
- Protein–protein interaction domains are therefore important not only for understanding individual proteins but also for understanding protein interaction networks. A single signaling protein can interact with several partners, and each of those partners can connect to additional proteins. The resulting network can contain many alternative routes for information flow and regulation.
- Disruption of these interactions can have major biological consequences. A mutation in an interaction domain can alter its ability to recognize a binding partner, while a mutation in the partner’s binding motif can have a similar effect. Such changes can disrupt signaling complexes, alter protein localization or modify the activity of downstream pathways. Importantly, a protein can retain its catalytic activity while losing an essential interaction, demonstrating that protein function depends on more than enzymatic activity alone.
- Aberrant protein–protein interactions can also contribute to disease mechanisms. Altered kinase activity can produce inappropriate phosphorylation and recruitment of SH2-containing proteins, while mutations affecting interaction motifs or scaffold proteins can alter the organization of signaling complexes. Similar principles apply to proteins involved in neuronal signaling, cell adhesion, cytoskeletal organization and intracellular trafficking.
- The modular architecture of SH2, SH3 and PDZ domains has also made them valuable tools in experimental biology. Researchers can isolate domains and binding peptides to study interaction specificity, engineer proteins containing selected interaction modules or use these domains to investigate signaling pathways. Interaction domains can therefore serve not only as subjects of study but also as molecular tools for manipulating protein networks.
- Several experimental techniques are used to characterize these interactions. Co-immunoprecipitation can determine whether proteins associate in cells, although it does not necessarily establish direct binding. Pull-down assays can investigate interactions under controlled biochemical conditions. Peptide-binding assays can examine recognition of specific motifs, while techniques such as surface plasmon resonance, isothermal titration calorimetry and microscale thermophoresis can quantify binding properties.
- Structural techniques can reveal how these domains recognize their partners at the molecular level. X-ray crystallography and NMR spectroscopy have been particularly important for understanding small interaction domains and their peptide-binding interfaces, while cryo-electron microscopy has become increasingly important for larger multiprotein complexes. Structural information can reveal which residues form the interaction surface and how mutations or modifications alter recognition.
- Bioinformatics can also help identify potential interaction domains and motifs. Conserved sequence regions can be compared with known domain families, while motif searches can identify potential binding sequences. However, motif identification alone does not prove that an interaction occurs in a living cell. Experimental validation is often required because short motifs can occur by chance and because interaction depends on cellular localization, protein abundance and regulatory state.
- The three domain families also illustrate different levels of molecular organization. An SH2 domain can recognize a phosphorylated residue within a signaling protein. An SH3 domain can recognize a proline-rich region. A PDZ domain can recognize a C-terminal sequence. These individual interactions can then bring proteins together into dimers, oligomers or larger signaling complexes. The complexes can participate in cellular pathways, and many such complexes can collectively form a protein–protein interaction network.
- This hierarchy connects naturally with quaternary structure. Quaternary structure describes how multiple protein subunits are organized into a functional assembly, whereas interaction domains describe some of the molecular mechanisms that allow those subunits to recognize and associate with one another. Not every protein–protein interaction produces a classical quaternary structure, because many interactions are transient or involve regulatory partners, but the two concepts are closely related.
- SH2, SH3 and PDZ domains also demonstrate why protein interactions should not be viewed simply as permanent physical attachments. Cellular protein complexes are continuously assembled, modified, redistributed and dismantled. Phosphorylation can recruit an SH2-containing protein, adaptor interactions can connect additional components through SH3 domains, and PDZ-mediated binding can organize proteins at a particular membrane. Changes in signaling state can subsequently reorganize the entire complex.
- The same principles apply to other interaction domains, including WW, PH, PTB, FHA and 14-3-3-associated recognition systems. Each domain family has its own structural characteristics and binding preferences, but together they form a large molecular vocabulary through which proteins recognize one another. Understanding individual domain families therefore provides a route toward understanding the much larger language of cellular protein interaction.
- From a broader perspective, SH2, SH3 and PDZ domains demonstrate how relatively small structural units can have effects extending across entire cellular pathways. A domain does not need to possess catalytic activity to be biologically important. By determining which proteins associate with one another, an interaction domain can control the location, timing, composition and activity of a molecular complex.
- The study of these domains therefore connects protein sequence, three-dimensional structure, post-translational modification and cellular signaling. SH2 domains translate phosphotyrosine signals into protein recruitment, SH3 domains connect proline-rich interaction motifs with signaling and assembly processes, and PDZ domains organize proteins through recognition of short C-terminal sequences. Together, these systems demonstrate how cells use modular protein–protein interaction domains to construct dynamic and highly regulated molecular networks.
- Understanding SH2, SH3 and PDZ domains also provides a foundation for exploring more complex concepts such as short linear motifs, intrinsically disordered regions, multidomain proteins, scaffold proteins and protein interaction networks. These topics extend the same central principle: biological information is encoded not only in catalytic sites and DNA-binding regions but also in the molecular surfaces through which proteins recognize, recruit and regulate one another. Protein–protein interaction domains are therefore essential components of the molecular architecture that connects individual proteins to functional cellular systems.
