This story originally ran on May 20.
Michael MacCoss, assistant professor at the University of Washington's Department of Genome Sciences, has begun a collaboration with Thermo Fisher Scientific to create a database of multiple-reaction monitoring mass spectroscopy assays for human proteins with the ultimate goal of developing assays for every protein in the human proteome.
The collaboration was inspired by the launch last October of Thermo Fisher's Pierce Human In Vitro Protein Expression kits, which MacCoss's lab will use to generate proteins to compile the database.
"I've been collaborating with Thermo on a number of projects," MacCoss told ProteoMonitor, "but this one kind of came out of the fact that Thermo has recently developed [the Pierce] in vitro translation kit that's based on a human cell lysate."
The MRM database project is similar to another MRM database effort that is also underway in Seattle by the Institute for Systems Biology in collaboration with Agilent and OriGene (PM 05/14/2010). MacCoss said he was unaware of the ISB project when he first proposed his work. He's since spoken to Robert Moritz, director of proteomics at ISB, and Eric Deutsch, senior database designer at ISB, about making his team's output available through the ISB's MRM Atlas, as well as on servers at the University of Washington.
"There will be complementary datasets available now," he said, noting that ISB will be using Agilent equipment and will therefore be collecting slightly different data. "So they will be complementary. They will collect data from a different protein source, and just like anything else, it's always good to do an experiment multiple times."
The Pierce kits that MacCoss will use for his project allow for expression of full-length proteins complete with post-translational modifications using human cell extracts containing all the necessary cellular machinery for protein expression.
MacCoss has just started work on the project, but, he said, initial experiments using the Pierce system have been promising.
"All the preliminary experiments look fantastic," he said. "It has very human-like modifications, which is not surprising since it is a human cell lysate. Glycosylated proteins are possible, and the expression levels appear to be very high."
The system is also relatively inexpensive. A typical Pierce kit costs around $10, MacCoss said, versus roughly $700 to order a protein ready-made from a company like OriGene.
"You need to do some cloning steps to get your plasmid, which takes some time and labor," he said, "but there are actually cDNA libraries within the community that are in vectors that are compatible with the Pierce in vitro translation system. It's very inexpensive to produce proteins as long as you have the plasmid."
MacCoss is currently in the process of ordering a library comprising around 7,500 protein clones. From there he plans to start producing proteins in 96-well-plate format, after which he'll begin mass spec analysis, which will also be done using Thermo Fisher equipment, specifically the Thermo Scientific TSQ Vantage.
Ultimately, MacCoss plans to encode the MRM dataset that the project produces into a document readable by Skyline, a free, open-source computer program that he's developed for building and analyzing SRM and MRM assays. Skyline, which MacCoss said is currently being used in over 1,500 institutions around the world, is designed to integrate spectrum data and assays from a variety of mass spec libraries and enables researchers to use them easily across platforms from a number of mass spec manufacturers, including Thermo Fisher, Agilent, Waters, and AB Sciex.
"We anticipate that a lot of people aren't going to access [the MRM data] through a web site," he said. "They're going to access it through a software tool that will directly interface with the instrument. If you're going to measure a hundred proteins, that's a lot of individual things to cut and paste. So what you'd want to do is to be able to just type in protein names and build an instrument method automatically based on those protein names or protein sequences."
MacCoss has been employing the system on a smaller scale as part of his work on the Verification Working Group at the National Cancer Institute's Clinical Proteomic Technology Assessment for Cancer network.
"[The CPTAC researchers] share Skyline documents because the Skyline documents are compatible regardless of the instrument vendor. With the Skyline document, Skyline can generate a method for all the individual vendor platforms and analyze the results directly from the RAW instrument data without file conversion," he said.
MacCoss anticipates that the mass spec analysis for the MRM database he and Thermo Fisher are working on will be done at a number of sites.
"I've been discussing with colleagues of mine who have mass spectrometers who are willing to help analyze portions of the protein sample that will be produced," he said. "We're planning on doing a portion here, some will probably be done at Thermo Fisher, some will probably be done at Pierce, and others will be done in other academic labs. It will be a very collaborative effort."
The precise terms of his collaboration with Thermo Fisher are still being determined, he said. In particular, the parties are still working out what sort of contribution of material resources Thermo Fisher might make. The company has committed to helping develop methodologies for the project, MacCoss said.
"We have colleagues there who are going to help us through the individual steps with some of the method development," he said. "They clearly have a huge amount of expertise in doing the in vitro translation."