Extracellular SH3 domain containing proteins--features of a new protein family

In the year 1994, the protein MIA (melanoma inhibitory activity) was found to be strongly expressed and secreted by malignant melanomas and subsequent studies revealed that MIA has an important function in melanoma development and invasion. Multidimensional NMR-spectroscopy and x-ray crystallography revealed that recombinant human MIA adopts a Src homology 3 (SH3) domain-like fold in solution, a structure with two perpendicular antiparallel three- and five-stranded beta-sheets. SH3 domains are protein modules that are found in many intracellular signalling proteins and mediate protein-protein interactions by binding to proline-rich peptide sequences. Unlike previously described protein structures with SH3 domain folds, MIA is a secreted single-domain protein of 12 kDa that contains an additional antiparallel beta-sheet and two disulfide bonds. Furthermore, the charge surrounding the canonical binding site differs from that of classical SH3 domains. The two disulfide bonds are crucial for correct folding and function as revealed by mutation analysis. Therefore, MIA appears to be the first extracellular protein adopting an SH3 domain-like fold. MIA was shown to interact with fibronectin, and MIA-interacting peptide ligands identified by phage display screening are similar to the consensus sequence of type III human fibronectin repeats, especially FN14. Interestingly, recent data revealed that MIA can also directly bind to integrin alpha 4 beta 1 and alpha 5 beta1 and that it modulates integrin activity negatively. These findings suggest an interesting role of the SH3-domain proteins in the extracellular compartment. Recently, MIA homologous proteins with a sequence identity of 44% and a sequence homology of approximately 80% were determined (TANGO, MIA-2, OTOR). This clearly suggests that this structural device is used more frequently, in processes ranging from developmental changes to the interference of cell attachment in the extracellular matrix. Detailed studies are necessary to determine the exact function of the MIA homologous proteins. It will be interesting to know whether additional protein families can be identified which are secreted and carry SH3 domain-like modules, in addition to elucidate what the specific physiological targets of this protein family are.