Proteins interact with each other to form complexes, and these complexes can be dynamic and interchanging as they relay biological signals. However, despite their central importance to biology and disease, protein complexes can be difficult to visualize and assess. There are even more technological barriers to analyzing protein complexes when they originate from non-genetically engineered human cells, such as those that would be provided in clinical patient samples. Currently, because physiologic protein complex profiles are virtually unobtainable, clinical practice cannot use them to assist in human health endeavors. We propose to advance a new strategy by mounting a new assay platform, MIF, to bring the analysis of physiologic, human protein complex profiles online. We have already learned how to overcome and control for the technical hurdles, leading us to launch MIF by targeting 21 human proteins that can participate in 231 inter-protein associations (Specific Aim 1). MIF will allow for analysis of small samples and high-throughput formatting, favoring its adoptability for primary human samples originating from clinical patients. Data analysis will involve the generation of Bioinformatics strategies (Specific Aim 2) to focus on three unique parameters of protein complexes: protein abundance, protein multiplicity in shared complexes (commonly thought of as protein plurization/aggregation), and heterotypic protein associations. We will field-test MIF by applying it to the analysis of human protein complexes that may be associated with the immune suppression that accompanies the universally lethal brain cancer, glioblastoma (Specific Aim 3). Together, MIF and its unique analysis will make available the acquisition of physiologic, human complex profiles. We propose that the glioblastoma- derived MIF data will exemplify a new strategy for analyzing these complexes, and illustrate its general applicability to many fields of study and classes of disease.
We will mount a new assay platform that will enable the concurrent assessment of large collections of physiologic, human protein complexes. The data generated by this methodology will be handled by novel Bioinformatics and Biostatistical analyses examining protein abundance, protein multiplicity in shared complexes, and heterotypic protein interactions. This method/analysis system will be field-tested in studies focusing on the universally lethal human cancer, glioblastoma, in search of protein complex profiles with predictive value for prognosis or therapeutic response.
