Modular protein domains mediate protein-protein interactions in various cellular proteins, including adaptors, enzymes and scaffold proteins. Src homology 2 (SH2)domains comprise a relatively large group of domains, which mediate inter- and intramolecular interactions by binding to tyrosine-phosphorylated ligands. Phosphorylation of tyrosine residues in proteins controls many facets of signaling in multicellular organisms, and increased tyrosine phosphorylation is associated with uncontrolled cell growth. Another group of domains, Src homology 3 (SH3) domains, mediates protein interactions by binding to ligands containing a polyproline type II (PPII) helix. SH3 domains mediate generally relatively low specificity and affinity interactions, although this might be due to the use of short ligands lacking the regions flanking the binding interface of SH3 domains. These outside regions are known to contribute to the ligand binding of SH3 domains. POSH proteins, which have been implicated as scaffold proteins in the JNK-mediated apoptosis signaling cascade, contain SH3 domains. POSH proteins also participate in protein trafficking and the regulation of protein degradation via ubiquitination.

The importance of tyrosine phosphorylation for many biological processes has led to an interest in profiling global tyrosine phosphorylation in cells. However, many of the methods utilized are either not suitable for comprehensive phosphotyrosine profiling or have technical limitations regarding their applicability for high-throughput analysis. In our study, we exploited the fact that the tyrosine phosphorylation of proteins creates binding sites for SH2 domains. A far-Western approach and a newly developed reversephase Rosette assay were used to profile the tyrosine phosphorylation of selected ligands (platelet endothelial cell adhesion molecule-1; PECAM-1 and p21-activated kinase 2; PAK2) as well to assess global tyrosine-phosphorylation levels from cell lysates. SH2 domains exhibited different protein binding preferences,which reflected in their ability to recognize a specific subset of cellular proteins. Our methods elucidated SH2 interactions at the protein level and demonstrated increased binding of SH2 domains to adherent cells in addition to adhesion-specific SH2/ligand interactions.

The human genome was found to contain 296 different SH3 domains, and these domains were used in a phage display approach to determine which SH3 domains bound most strongly to our target proteins of interest: Nef, PAK2 and a disintegrin and metalloprotease 15 (ADAM15). An unbiased system for simultaneously assaying the complete human SH3 proteome was used to determine the preferred SH3/ligand interactions; this approach provided valuable information without the limitations caused by short peptide ligands or the skewing of variables caused by more indirect methods. Our approach identified both previously reported and novel SH3 domains that were capable of binding to the target proteins with nanomolar affinities. In addition to providing information regarding the SH3-mediated binding kinetics, this method also identified novel signaling proteins, such as the PAK2-binding scaffold protein plenty of SH3 domains 2 (POSH2).

POSH2 was found to be a highly homologous new member of the previously identified POSH family of proteins. POSH2 was shown to contain four SH3 domains and a RING domain, which provides ubiquitin E3 ligase activity to the protein. Activated Rac1, a GTPase, was shown to interact with POSH2, and this interaction was mediated by the partial Cdc42/Rac1 interactive binding (CRIB) domain. Moreover, the interaction of POSH2 and Rac1 suggests that POSH2 acts downstream of Rac1 in JNK-mediated apoptosis.

In conclusion, our results from the SH2 domain study provided information regarding likely in vivo interactions and changes in the concentration of SH2 binding sites under various conditions. Deciphering the global phosphotyrosine pattern and identifying activated signaling pathways are also important for understanding aberrant cellular functions. Thus, global phosphotyrosine profiling and quantification by SH2 domains could be used as a diagnostic tool in clinical applications. SH3 domains were suggested to have a more prominent role in mediating cellular interactions, and these domains bind with higher specificity than was previously anticipated. Moreover, the SH3 phage library system proved to be a valuable tool for deciphering the wiring of SH3-dependent signaling networks within the cell. POSH2 contains an SH3 domain and is the third member of the POSH protein family. These POSH proteins contain a highly homologous domain structure consisting of a RING finger domain, multiple SH3 domains and a partial CRIB domain, which is crucial for binding to Rac1.