Christoph Bock, PhD, a principal investigator at the CeMM Research Center for Molecular Medicine in Vienna, Austria, talks to contributing editor Tanuja Koppal, PhD, about the new CROP-seq technology for single-cell CRISPR sequencing that his group has developed and its potential uses for functional screening. He talks about some of the inherent challenges working with CRISPR and other phenotypic screens and how some of this can be overcome with new approaches like CROP-seq.
Q: Can you share with our readers the recent advances in Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) screening and some new applications that are being developed?
A: CRISPR technology has revolutionized the way scientists investigate the biological function of genes. It is now possible to edit or delete genes much faster and more efficiently with CRISPR than with any alternative technology. In fact, the method works so well that it is possible to add CRISPR guide-RNAs (i.e., the short RNAs that direct the CRISPR/ Cas9 genome editing to specific genes) against thousands of genes to a large pool of cells, and use Darwinian selection to find those guide-RNAs that target genes that make the cells grow faster or slower, or affect their susceptibility to a drug.
This approach is called pooled CRISPR screening. Typically, you first infect a large number of cells with guide-RNAs that target many (or all) genes in the genome. Then you apply the drug or the virus, wait until the nonresistant cells die, and you sequence the guide-RNAs of the surviving cells. Pooled CRISPR screening effectively pinpoints genes involved in the cell’s sensitivity to a drug or a virus.
However, this approach does not work so well when the phenotype of interest is more complex than just counting surviving cells. In biomedical research, you may, for example, be interested in upregulation of a key cancer pathway, dedifferentiation to a more immature state, or metastatic potential. These cellular phenotypes are not readily accessible to classical CRISPR screens, but they can be inferred from a cell’s transcriptome. This is why we developed the CRISPR Droplet Sequencing (CROP-seq) method for pooled CRISPR screening with single-cell transcriptome readout. In other words, for each guide-RNA, we can measure how the inactivation of its target gene influences the transcription of other genes in the genome, which is highly informative for understanding the biological function of the target gene. Importantly, with CROP-seq, this can be done in high throughput, studying thousands of genes and hundreds of thousands of cells in parallel.
Q: Can you explain how CROPseq works and elaborate on some of its inherent limitations, as well as its promise?
A: When we study the behavior of cells, transcriptome profiling by RNA-seq is by far the most informative assay—it gives us the expression levels of thousands of genes and a deep insight into the biology of the cells under investigation. However, combining CRISPR genome editing with RNA-seq was labor-intensive and low-throughput because you need a separate cell culture dish (or a well on a 96-well plate) for each gene that is being investigated, in order to keep the cells that are infected with different guide-RNAs separated from each other. The fundamental idea behind CROP-seq and single- cell CRISPR sequencing is that this separation is no longer necessary when performing single-cell RNA-seq—in this case, the cell membrane of each individual cell provides the separation between different guide-RNAs and the changes in gene expression that are caused by the deletion of each targeted gene.
Exploiting this novel combination of CRISPR genome editing and single-cell sequencing technology, CROP-seq (http://www.nature.com/nmeth/journal/v14/n3/full/nmeth.4177.html) makes it possible to perform a CRISPR screen for anything that can be inferred from the transcriptome. For example, there are gene signatures for the activity of signaling pathways, metabolic states, drug resistance, and many other aspects of cell biology. Indeed, if you ask scientists, “Assume that there’s only one assay that you could run on your cells, what would give you the most information?” I think most will go for transcriptome profiling/RNA-seq as their first choice. Using CROP-seq, we can search for genes involved in cancer pathways, dedifferentiation, and metastatic potential—or anything else that is reflected in the transcriptome of the cell. An additional advantage is that you do not need to know upfront what you are looking for. Based on the same CROP-seq dataset, you can run screens for many different pathway signatures purely on the computer.
Practically speaking, a CROP-seq screen works as follows:
- Select an interesting cellular model (e.g., a cell line in which you induce or knock down a cancer gene).
- Design a guide-RNA library for genes of interest (our recent paper showed a proof-of-concept with more than 100 guide-RNAs, but CROP-seq is limited only by the cost of single-cell RNA-seq, and libraries targeting thousands of genes are entirely feasible and increasingly affordable).
- Take the cells, infect them with the guide-RNA library, and induce any biological stimulus that may be of interest (e.g., treatment with a drug or infection with a pathogen).
- Perform single-cell RNA-seq before and after applying the stimulus.
- Perform bioinformatic analysis to link guide-RNA target genes to their transcriptome responses.
Q: How does your recent work in developing CROP-seq help with driving the field forward in terms of serving new applications and industries?
A: It will be very exciting to apply CROP-seq technology to complex and heterogeneous tissue, including primary tumors where each cell type may respond differently to a drug administered to the tumor as a whole. Led by my CeMM colleague Stefan Kubicek, we have previously shown that single-cell RNA-seq can be used to investigate celltype specific response to drugs directly in primary human tissue (human pancreatic islets treated with an anti-malaria drug to induce insulin production, recently published in Cell: http://www.cell.com/cell/comments/S0092-8674(16)31531-8).
With CROP-seq, we can now test whether different cell types within the same tumor depend on different genes in the way they handle, for example, the challenges of chemotherapy. More generally, given that CRISPR screens have become the method of choice for discovering new biology in a broad range of diseases and other applications, CROP-seq has huge potential to add deep regulatory information to any CRISPR screen, shortcutting much of the—typically laborious and time-consuming—validation and functional workup that follows a classical CRISPR screen focusing on cell survival.
Q: What would you advise researchers looking to invest in new CRISPR and genomic technologies?
A: Just give it a try—we have published a detailed open-source protocol for CROP-seq on the website (http://cropseq.computational-epigenetics.org/) and the plasmid is available via AddGene (https://www.addgene.org/86708/). Indeed, approximately 100 laboratories from a broad range of fields have already obtained the CROP-seq plasmid and are currently exploring how CROP-seq can advance their research.
Christoph Bock is a principal investigator at the CeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences. His research focuses on dissecting the role of epigenetics in cancer and on developing high-throughput technologies for precision medicine. He is also a guest professor at the Medical University of Vienna, scientific coordinator of the Biomedical Sequencing Facility at CeMM, and adjunct group leader for bioinformatics at the Max Planck Institute for Informatics. He has received several research awards, including the Max Planck Society’s Otto Hahn Medal (2009), an ERC Starting Grant (2016-2021), and the Overton Prize of the International Society of Computational Biology (2017).