With more than 1 trillion neurons in the human brain and millions of proteins in each neuron, getting a handle on exactly how all these proteins work together is a Sisyphean task. But mapping out protein interactions in neurons is precisely what a group of researchers armed with a two-year, $2.6 million grant from the National Institutes of Health is attempting to do. Researchers from the University of Miami are using the grant to conduct a systematic survey of "isPIN," or in situ protein-protein interactions, with the long-term goal of reconstructing genome-wide protein-protein interaction networks. "The map now reflects the real-life situation for the proteins. People need to understand that, previous to this work, the protein-protein interaction map was based on either processed cells — dissected and processed, homogenized and extracted protein — or the highly controlled, but artificial, situation where two fluorescent proteins co-exist in each cell and then assay the interaction," says Akira Chiba, the biology professor at the University of Miami charged with leading the project. "This is the first time that endogenous proteins are studied in a natural situation. We are watching the Drosophila protein in the neuron in undissected animal, so that 'inductness' of the environment that surrounds the protein, that is new."
Chiba says that only now, with the recent advancements in genomics, imaging, and high-performance computation, could a project of such massive scale be attempted. "The new microscope design allows us to collect the images on almost two orders of magnitude faster than previous technology. The same thing could have been, but it probably would have taken 100 times longer. Time is the breakthrough," he says. "Other things have improved on a smaller scale, but still you add them together and everything works faster and more efficiently than before."
Furthermore, these faster and more efficient technologies will provide a lot of data. "We anticipate the amount of the raw data [will be] massive, easily surpassing the amount of raw data that the genome project handles, and for this part of the project, which is less than 0.1 percent [of what can be mapped] ... we're approaching terabytes," Chiba says.