New technology puts a spatial lens on CRISPR screening

Recently, scientists have been able to explore gene circuitry in individual cells using methods that suppress particular genes and measure the impact on the expression of other genes. These methods, however, fail to capture spatial information such as the effects from, or on, neighboring cells, which can provide important clues to a cell or gene’s role in health and disease.

Now, a technology developed in the Spatial Technology Platform at the Broad Institute of MIT and Harvard builds on these methods with cutting-edge spatial advances. Their method, known as Perturb-FISH, combines imaging-based spatial transcriptomic measurements with large-scale detection of CRISPR guide RNAs. The researchers demonstrated Perturb-seq’s ability to uncover new cellular and functional insights, including the effects of autism-related genes on cellular activity and the interactions between human tumor and immune cells in an animal model. Further refinements to Perturb-FISH could make it even more widely accessible and enable a range of new biological investigations. The technology and demonstrations of its utility are described in Cell.

“With Perturb-FISH, we’ve developed a powerful new way to examine the roles of genes and genetic circuits in tissue development, homeostasis, and dysfunction,” said co-senior author Sami Farhi, director of the Spatial Technology Platform at the Broad Institute. “Our team is dedicated to devising spatial tools for the benefit of the scientific community, and we hope this new method is just the first of many more that we’ll build and share.” Farhi led the work along with co-senior author Brian Cleary, a former Broad Fellow and Merkin Institute Fellow who is now an assistant professor at Boston University.

Gone FISH-ing

Previously, researchers had combined single-cell RNA sequencing with CRISPR screening to examine gene networks within cells using tools such as Perturb-seq, but the cells’ spatial context wasn’t captured. Another method known as optical pooled screening measures the effects of gene editing perturbations on cell fitness or other phenotypes, but doesn’t capture gene transcriptional states.

Scientists in the Spatial Technology Platform aimed to develop a comprehensive method that could measure at once which genes are altered in a cell, which genetic perturbation caused the changes, and the location of those affected cells in relation to other cells.

Key developments that made Perturb-FISH possible were computational methods developed by Cleary and a new way of amplifying the signal from single molecules (either CRISPR guide RNAs or gene transcripts) so they can be detected over background levels of fluorescence. Traditional amplification methods don’t work with molecules as small as guide RNAs, so first author and former postdoctoral researcher Loϊc Binan devised an innovative strategy to generate many local copies of each guide RNA at its original site. By combining that with a fluorescence-based spatial transcriptomic method called MERFISH, Perturb-FISH can reveal both the identity of each perturbation and the cell’s transcriptome in their spatial context.

The researchers demonstrated Perturb-FISH’s ability to measure the effects of 35 genes that help regulate immune cell response to stimulation, finding that their results aligned with similar data generated using Perturb-seq. They also showed how it can shed light on intercellular networks, finding that cell density can alter the effects of gene perturbations and that non-perturbed cells can be impacted by perturbations in neighboring cells.

To illustrate Perturb-FISH’s potential for investigating functional effects of even subtle genetic adjustments, the researchers used CRISPR inhibition to turn down genes associated with autism in human brain cells known as astrocytes and observed changes in calcium activity and gene expression.

In addition, the research team tested the method in a more complex system, using a xenograft model developed in the Broad Cancer Program. After perturbing genes related to the NF-kB pathway in human tumor cells, they used Perturb-FISH to measure the genetic interactions between tumor cells and responding immune cells in the animal model. “We were excited to see that we could observe effects in a more complicated system with multiple cell types interacting with each other and still pick up interesting biology,” said Farhi.

Members of the platform are now working to make Perturb-FISH more robust and more accessible, for example, by modifying the method so that only a single imaging instrument is required to run the method. They also want to increase the number of genes they can examine in each experiment, which is currently limited to 500. “Because we developed this in the Spatial Technology Platform, it allows us to take what would typically just be an academic project and keep working on it and refining it to the point that it can be more broadly used,” said Farhi.

Along with related methods recently launched by other research groups, Perturb-FISH represents a step forward in the spatial technology field that will give users more experimental opportunities and choices. “We’re happy to see others pursuing these technologies, which bring exciting possibilities for scientists wanting to launch spatial investigations,” said Farhi. “The more methods we all have, the more likely people are to use them, which is our ultimate goal.”

Other Broad scientists contributing to this work include Aiping Jiang, Serwah Danquah, Vera Valakh, Brooke Simonton, Jon Bezney, Robert Manguso, Kathleen Yates, and Ralda Nehme.