NEW YORK (GenomeWeb) – Researchers at the Netherlands Cancer Institute (NKI) have developed a technique that combines antibody-based protein detection with cell sorting and sequencing to study cellular processes.
In a paper published this week in Nature, the scientists described how they used the approach to investigate various cell functions, including transcriptional induction, regulation of protein abundance and splicing, and protein signaling pathways including MAPK, AKT, and WNT. Thijn Brummelkamp, an NKI professor and senior author on the paper also said he has co-founded a company, Scenic Biotech, that plans to use the approach to aid drug development.
The method begins with random mutagenesis of human haploid cells. These cells are then stained with an antibody to the protein target of interest, and then sorted using fluorescence-activated cell sorting into two populations, one corresponding to cells with the highest abundance of the target protein and the other to cell with the lowest abundance of the protein.
These cells are then sequenced to detect mutations created by the initial mutagenesis, and these mutations are analyzed for how they did or didn't affect the abundance of the target protein.
"Basically, every individual cell we have selected for a certain phenotype yields a mutation that we can assign back to the genome," Brummelkamp said. “We select millions of cells for certain proteins states and subsequently measure mutations in all human genes in those cells. Mutations in genes that influence this cellular trait will either be enriched or depleted in this selected cell population in contrast to mutations in genes that do not affect the cellular phenotype.”
By identifying these genes that are enriched or deleted for mutations, Brummelkamp and colleagues including first author Markus Brockmann hope to gather data on genes that control specific proteins that will subsequently enable them to better understand the underlying gene networks and their function.
In the Nature study the researchers used for a proof of concept the induction of interferon-regulatory factor 1 (IRF1) by interferon-γ, which, they noted, is a system known to be under strong genetic control. Using the approach to analyze mutated cells stained for IRF1, they identified as enriched all previously described components of the network as well as a number of additional regulators, indicating the ability of the technique to, for some targets at least, provide a comprehensive overview of the involved pathways.
Developing a streamlined workflow capable of sorting and sequencing the large numbers of cells was a technical bottleneck, Brummelkamp said, noting that in a given experiment they are looking at on the order of tens to hundreds of millions of cells. Once in place, though, the method provides a relatively straightforward approach to studying cellular events on a large scale, he said. “We stain cells with an antibody and use DNA sequencing as readout, and out of that you basically get cell biology. In some cases complete singling pathways"
A typical experiment, he said, costs between €4,000 and €5,000 ($4,500 to $5,600) to run.
One notable finding from the study was the identification of what appears to be a gene involved in G-protein coupled receptor signaling and AKT signaling.
"The AKT pathway has been extremely well studied, and, indeed, we found many of the components that other people have previously identified," Brummelkamp said. "But to our surprise the strongest outlier [in their AKT analysis] was a gene that had never been linked to AKT signaling before, and it turned out to be an off switch at the level of GPCR signaling. So that is an example where in a very well-studied process we were able to identify the known players but also this new one."
Brummelkamp said that his lab plans to apply the technique to a large number of important protein targets to build a genetic wiring map linked to cell biological processes.
"There are many ways people have done this—using large-scale protein-protein interaction data or transcriptional profiles- and using that to build regulatory networks," he said. "What we here use are quantitative protein phenotypes, so we can basically measure the state of a protein and link the [genetic] regulation back to this."
In the Nature study the researchers demonstrated their method for 10 ten different protein states, but Brummelkamp noted they hope to eventually use it for hundreds of proteins. "In doing so we will get a very good idea of how the cell is working and how the genetic wiring looks like," he said. "This would assign new functions to genes in different processes and visualize how processes talk to each other, so that is one main goal."
The researchers also have in mind variations on the experiments presented in the recent study, looking, for instance, at normal cells and cells harboring a specific mutation.
"By doing that we can identify how the genetic regulatory network changes if you take out one player," Brummelkamp said. "Are there regulators coming in? Are there others falling out? That is the kind of game we can play, and by doing that we can decipher how genes work together, so called genetic interactions such as genetic suppression mechanisms, for example — what kind of mechanisms step in to cause a phenotype when you alter a gene?"
The question of suppression mechanisms will be central for Amsterdam-based Scenic Biotech, which Brummelkamp launched this year as a spin out of NKI and Oxford University and which will use his lab's technique to study the genetic networks underlying various diseases and develop new therapeutics.