This story originally ran on Dec. 3.
By Tony Fong
Attempting the "first systematic inspection of when and where each protein-protein interaction takes place in vivo," investigators at the University of Miami, Coral Gables, have set on creating a map of protein-protein interactions in neurons.
The work, being funded with a two-year, $2.6 million stimulus grant from the National Institutes of Health, is aimed at using imaging technologies to create an in situ protein-protein map of Drosophila with an ultimate goal of reconstructing "genome-wide protein-protein interaction networks within each and every subcellular compartment of neurons at progressive steps of their development," according to the grant abstract.
In it, the researchers said they would isolate green fluorescent protein trap lines and create transgenic lines for "neurologically 'relevant' proteins tagged with either GFP or monomeric red fluorescent proteins." Crosses will be used to combine the green and red tagged proteins in a single fly, and the researchers will then determine the localization of each protein "and reveal the dynamic interactions between proteins in cell body, axon, dendrite and/or synapse of intact neurons within a whole organism during development."
The research will target 100 Drosophila proteins with homologs to human neuron proteins and include synaptic proteins and axodendritic proteins. The 100 proteins make up about 10,000 protein-protein pairs, said Akira Chiba, a professor of biology at UM and the principal investigator on the project.
The map he and his colleagues intend to create would have a "specific context such as axonal subdomain versus dendritic subdomain, or …axon terminal before synaptogenesis versus axon terminal before synaptogenesis," he told ProteoMonitor this week.
"We predict, or we hope to see, that proteomic maps even in the same neuron — depending on whether you look at a dendritic compartment or axonal compartment, or before synaptogenesis or after synaptogenesis — will change," Chiba said.
As part of the project, Chiba and his colleagues plan to generate 500 transgenic Drosophila lines, and analyze more than 1 million 3D images occupying 1 terabyte of memory space.
A web page describing the project is under development and is expected to go live later this month at http://www.miami.edu/iSPIN. Data from the project will also be deposited on the site.
The research is being conducted within the context of neurons under normal conditions, but Chiba said the research can be adapted for studying protein-protein interactions under disease conditions.
While protein-protein interactions are being investigated by a number of researchers and labs throughout the world, the current project is "a new way of looking at cell biology in terms of protein interactions," according to Chiba.
"Instead of knowing or finding interactions, we will actually begin, for the first time, to know what actual interactions take place within the neuron," he said. Current research in protein-protein interactions has depended largely on yeast two-hybrids, pull-downs of protein complexes, and luminescence-based mammalian interactome technology.
Those approaches can "tell us what could be potentially happening among proteins — the maximum possible interaction map," Chiba said. But the interactions that actually happen in a specific cellular context "is part of the unknown. … And then the changes in actively engaged interactions, I think, will tell us how different cell types actually achieve different cellular molecular profiles," he added.
In their abstract, the researchers wrote that their project would provide "a much-needed intellectual infrastructure to transform definitions of the molecular circuitries affected by neurological disorders as well as strategies to repair/protect neurons."
Key to the work is a $1 million custom-designed microscope developed by a former student of Chiba's, Daichi Kamiyama, now a co-investigator on the current project. Denver firm Intelligent Imaging Innovations builds and commercializes the system.
[ pagebreak ]
The instrument allows for quantitative 3D FRET detection. While FRET quantitation based on lifetime imaging — which is based on differences in the exponential decay rate of the fluorescence from a fluorescent sample — has been around for several years, the breakthrough capability of the Kamiyama-designed system is the speed, between 50 and 100 times faster than other commercially available multi-photon imaging systems, Chiba said.
"The speed allowed us to start thinking about potentially scaling things up and looking at many different protein pairs in relatively short period[s of time]," he said.
In their original grant application, Chiba and his colleagues requested nearly $5 million — about half of what they eventually received — for their project with the aim of investigating up to 1,000 proteins making up 100,000 protein pairs.
The researchers said in their grant abstract that the project would deliver not only a context-rich proteomics resource but also a "new intellectual infrastructure for determining the molecular circuitries affected by neurological disorders, aging, or drug addiction, and designing strategies to repair and/or protect neurons."
But though the project is now one-tenth of the originally envisioned plan, the goal hasn't changed, Chiba said.
"At the end of this two-year project … we [will be] nowhere near completing the entire proteome," he said. "We just want to have enough progress to convince [others] that this is feasible and we want to be ready to request subsequent funding."
Cutting their numbers down to 100 proteins has resulted in "lively discussion ... and a lot of brainstorming," Chiba said. As they continue choosing their proteins to investigate, they want to make sure that they have enough positive controls, "so we pick the set of signaling pathway proteins that are known outside of cells, or in yeast, [to] bind to each other. These are the ones we think will work as a positive control."
They also want negative controls — proteins that the researchers do not believe "FRET with each other." Also included are molecules that are known to be neurons but are suspected to work only in dendrites but not in axons, or vice versa.
"Those are the ones [that will be] included to show that our imaging [allows] us to separate FRET interactions in a different subcompartment of the neuron," Chiba said.
Lastly, they plan to include some proteins that have been implicated in neuronal and mental illnesses.
Also collaborating on the project is Susan Celniker, co-director of the Berkeley Drosophila Sequencing Program at Lawrence Berkeley National Laboratory. Celniker, who devised and is providing the reagents for the project, is also collaborating with the UM team to generate most of the constructs for creating transgenic fly lines for the project.
Chiba and his team are currently preparing the Drosophila for the research, cloning the genes, and modifying them with transformation vectors "so that we can produce new transgenic lines of Drosophila stocks."
Those stocks will then be crossed to cell-specific drivers. That will take a few months to be completed, and afterward "we anticipate a lot of intense high-volume image processing to make sense of the massive data."