Immunoglobulin superfamily (IgSF) proteins play important roles in protecting the human body from infectious diseases and tumorigenesis; on the other hand, their malfunction can lead to automimmune diseases. Because IgSF proteins function in immunity by specific trans-cellular noncovalent interactions between antigen-presenting cells and T cells, a molecular-level understanding of IgSF:IgSF binding interfaces would be of great aid to the design of novel immunomodulatory therapeutics. Excluding antibodies, the human proteome currently contains 477 extracellular IgSF proteins, of which only a quarter have documented binding partners. Given the volume of unexplored extracellular IgSF:IgSF interactions, a purely wet-lab approach to completing the IgSF interactome-the network of all known IgSF:IgSF interactions-would be prohibitively expensive. On the other hand, current computational molecular interaction prediction approaches are unsuitable for interactome prediction as they are either computationally intractable when attempted on large molecules such as proteins due to their inability to sample the entire conformational space or produce inaccurate results due to their inability to distinguish binding from non-binding protein pairs. Our goal is to develop a computational method that can be used to identify interacting IgSF receptor-ligand pairs. To accomplish this goal, we will first combine structural similarity-based and sequence-based approaches along with hidden Markov model profile-based functional sub-classification of the IgSF to identify the binding interfaces of IgSF proteins. Next using molecular dynamics simulations, we will sample the potential energy landscape of target receptor IgSF protein binding interfaces and design an optimal complementary ligand protein interface, which will then be evaluated to fit existing IgSF proteins. We hypothesize that each receptor interface can be characterized by a unique spatial fingerprint-an extended pharmacophore which we will call the residue-specific functional atom field (rsFAF)-which represents the energetically favorable positions of key functional atoms and can be used to identify cognate ligands. Our methods will be validated using a test set of known IgSF:IgSF complexes with available crystallographic structures.
Immunoglobulin superfamily (IgSF) proteins are crucial to a great variety of biological processes, including the control of innate and adaptive immunity and the suppression of tumorigenesis; in immunity, IgSF proteins interact at the cellular interface between antigen-presenting cells and T cells known as the immunological synapse. Our goal is to develop a computational method to predict the interactions between all IgSF proteins- or at least to shortlist candidate interactions to an experimentally manageable few-by gaining a molecular- level understanding of the immunological synapse. The impact of our research will be twofold: not only will we identify unknown interactions between IgSF proteins that could be potential immunomodulatory drug targets, but we will also elucidate a molecular-level understanding of the immunological synapse that could serve as a scaffold for the process of immunomodulatory drug design itself.
