Proteins in the tumor necrosis factor receptor superfamilies'(TNFSF/TNFRSF) are expressed on cells of the immune system, as well as on non-lymphoid structural cells such as stromal, epithelial, and endothelial cells. They have a broad range of activities in regulating cell-cell communication in many organs. TNF family molecules have canonical TNF homology domains with ?-sandwich jelly-roll structures, and TNFR family molecules contain variable numbers of ligand-binding extracellular cysteine-rich domains (CRDs). TNFSF are active largely as cell-surface or soluble homotrimers, and the crystallized complexes of 11 interactions solved to date show that the trimers engage 3 individual TNFRSF monomers. Whether a single trimeric complex is sufficient for cellular activation is not known, but oligomers of the basic trimer may also need to form for maximal signaling. How each TNFRSF forms a functional signaling unit likely differs due to variations in the ligand structures as well as how they are recognized by the receptors. TNFSF trimers have been subdivided into 3 groups: conventional/bell-shaped, EF-disulfide bond linked/globular, and divergent/blooming flower-like. These structures have been suggested to regulate the spacing between TNFRSF monomers when binding their ligands on the cell membrane which has been theorized to control the intracellular recruitment of adaptors that link to signal transduction pathways. We have studied 4-1BBL (CD137L/TNFSF9), which was recently crystallized and shown to exert the most open structure of any of the divergent/blooming flower-like TNFSF molecules. This suggests it may not be able to cluster 3 monomeric 4-1BB (CD137/TNFRSF9) molecules as effectively as other TNFSF proteins, or may not efficiently allow oligomers to form. Therefore 4-1BBL might need additional co-receptor proteins to engage 4-1BB to aggregate into a fully functional signaling unit. We have now discovered that 4-1BB interacts with Galectin-9 (Gal-9), a member of a family of animal lectins that interact with proteins in part via carbohydrates. In functional studies of immune cells, and in disease models in vivo, we have found that Gal-9 is required for the activity of 4-1BB. We hypothesize that Gal-9 is a co-receptor protein for 4-1BB that enables 4-1BB monomers to complex into a functional signaling unit. We will determine the three dimensional structures of human 4-1BB complexed with Gal-9, and of 4-1BB complexed with 4-1BBL, using X-ray crystallography, and also attempt to crystallize a ternary complex of all three molecules. These studies will define a new paradigm for structural interaction in the TNFR family.
Tumor Necrosis Factor (TNF)-related cytokines provide key cellular communication systems in mammals. Inhibitors of TNF are successful protein-based drugs for controlling inflammation in autoimmune diseases, and reagents to many of the molecules related to TNF are in pre-clinical or clinical development. There is much still to learn about how TNF family molecules engage their receptors to form functional cell-signaling units. We will characterize and crystallize novel complexes of a TNF family member with an unrelated protein that we have shown is essential for regulating cellular activity. These studies will define a new paradigm for structural interaction in the TNF family and provide important insights into the function of a member of this protein family that could be relevant to human clinical intervention.