NEW YORK (GenomeWeb) – An understanding of cell function is a necessary and basic part of life science research that can lead to a greater understanding of both healthy and diseased biological states. But the tools for efficiently conducting functional genomics studies are still a little lacking.
Paul Blainey — a core member at the Broad Institute and an assistant professor in MIT's department of biological engineering — recently received a four-year grant from the National Human Genome Research Institute for a project that aims to fill some of the gaps.
According to the grant's abstract, pooled methods for genome-wide screening currently require that cells are selected based on growth advantage or physical purification. Previous studies involving image-based, high-content screens using overexpression and RNA interference (RNAi) have uncovered novel genes involved in complex phenotypes, but conducting such microplate-based screens of clonal cell populations at the genomic scale is expensive, labor intensive, and requires specific automation expertise. Other approaches such as single-cell transcriptomics, however, can't access a wide enough range of complex and dynamic disease-associated phenotypes.
Blainey and his team proposed, therefore, to develop a new type of genomic perturbation and screening system that would combine the advantages of pooled perturbation with imaging assays for single-cell arrayed readout of complex phenotypes. But they also took it a step further, introducing a new technology that wasn't available a few years ago: CRISPR genome editing.
"Specifically, we will screen pooled genomic perturbations (with barcoded CRISPR-Cas9 single guide RNAs) using microscopy to read out phenotypes and to identify perturbed genes at the single-cell level," they wrote in the abstract. "Perturbed genes will be identified by sgRNA-associated expressed barcodes in situ [and read by] RNA fluorescence in situ hybridization (FISH) or in situ sequencing (IS)."
Blainey and his team will receive $1.1 million in 2017 for the project, including about $750,000 in direct costs and $370,000 in indirect costs. They'll receive about $700,000 in direct costs every subsequent year for the duration of the grant, he told GenomeWeb.
"The project originated when we had the mandate to take advantage of all these new CRISPR technologies that were coming out, especially with the focus on mammalian genomics here at the Broad, and to try to build new assays and rethink existing assays, especially considering the new angles you could take with CRISPR," David Feldman, a PhD student in physics at MIT, jointly advised by Blainey and Feng Zhang, said. "We chose [to tackle the challenge of] scalably collecting high-content information on genetic perturbations for modeling fundamental [biological] processes."
There are many kinds of information to be gleaned from functional genomic screens, Feldman added, but the depth of information that results depends on the type of screen that's conducted. A study done with RNAi reagents or a big image-based screen can yield spatial and temporal information about cells, live-cell imaging, subcellular localization, and potentially cell-cell interactions. Functional genomics screens that use the survival of cells or fluorescent-based sorting of cells as a readout of function are easier to do but don't yield that sort of data.
But the approaches that yield more information are also more challenging to conduct. "Basically, all these papers [that used RNAi for arrayed high-content screens] came from big labs with major investments in automation. If you look at the papers and pencil out how many microwells of cell culture they did for the assays, it's in the hundreds of thousands," Feldman said. "That's very different from the pooled screens which are a lot simpler. You have to maintain 10 or 20 flasks of cells for a couple of weeks. It's something that a single person can do without more than a standard training in tissue culture. So basically, what we want to do is combine the benefits of the two approaches, to have the much higher information content of imaging, with the relative ease and lack of robotics of pooled screening."
The researchers will then use CRISPR genome editing to perturb the cells in various ways and see how the changes affect function. Right now, they're planning to use the Streptococcus pyogenes Cas9 (SpCas9) nuclease to conduct loss-of-function screening, but Feldman noted that one benefit of this pooled screening approach is that it's compatible with all the CRISPR systems that are currently in use, as well as RNAi, open reading frames, and other editing methods. "With a slight tweak in the guide design and no significant changes to the pooled cloning procedures, the same libraries could be used for CRISPR interference and CRISPR activation," he said. "Newer CRISPR systems are also compatible."
Further, the team is planning to use an in situ sequencing method developed by Mats Nilsson and his team at Stockholm University in 2013 to read the perturbed genes in the imaging dish. Because Blainey and his colleagues are designing their own sgRNAs, they're able to add a barcode that will allow them to optimize Nilsson's sequencing method.
Avtar Singh, a postdoc who joined Blainey's group at the end of February, said adding the barcodes and reading them optically was "the missing piece" to other genomic function screening approaches, and something that's not possible to do with traditional next-generation sequencing.
"In the past, [we] read out the guide or the barcode by sequencing, and we need to be able to read that out in the imaging dish, in situ," Singh said. To do that, the researchers take the barcodes and amplify them using a technique called rolling circle amplification.
"Now you have thousands and thousands of copies of that one barcode at one spot," he added. "This is a way to convert those sequence of bases into a sequence of colors and then read out those colors over cycles of imaging. The color of that oligo will tell you the base information at that one position."
At the end of this project, Blainey and his team believe that both the data they produce and the tools they create can be useful in a number of contexts. First and foremost is the basic need to have a robust understanding of basic biology and how cells function absent any disease. For that, the team is collaborating with Nir Hacohen, a member of the Broad and director of the MGH Center for Cancer Immunotherapy; Iain Cheeseman, a member of the Whitehead Institute and associate professor of biology at MIT; and Arjun Raj, assistant professor of bioengineering at the University of Pennsylvania Perelman School of Medicine.
Blainey's team will also work with Jeff Cottrell, a director of translational research at the Stanley Center for Psychiatric Research at the Broad, who is looking at primary neuron cultures to classify and begin to understand GWAS hits in schizophrenia and other psychiatric diseases.
The data the team produces will get reported in publications, and the Broad will make the raw data accessible to the research community, Blainey said, adding, "We're developing some pooled cloning methods that we think will be robust and reliable for producing the specialized plasmid libraries and viral libraries to create the cell populations used in the screen. And then we're also working on fine-tuning the in situ sequencing protocol for reading the barcode on those perturbation reagents. We think all these things will be pretty robust."