At the DRSC-BTRR, we've been doing a lot of pooled-format CRISPR knockout screens in Drosophila cells. We're finding the results to be robust and reproducible. And best of all, the results have been informative, providing insights into diverse areas of biology.
Thinking about how to do CRISPR knockout screens in cells is a little different from thinking about how to do a genetic or RNAi screen in vivo or doing an arrayed-format RNAi screen. Read on for tips on how to design and carry out a screen. Notably, pooled-format screens do not require specialized equipment. You should be able to do a screen in your own lab. For Drosophila cell screens, we've made the 'screen-ready' cells available at DGRC (cell line #268) and the sgRNAs libraries needed for this approach at Addgene (134582, 134583, 134584). In addition, we're here to provide advice and other support along the way. Reagents are also available for mosquito cell screens.
Are you more of a visual learner? Click the attachment below to view an overall screen workflow.
Are you an 'okay but how much will it cost the lab to do it' type? To start, it's about 10x less costly than an arrayed-format cell screen. Please contact us to learn what to expect for costs if you decide to do a screen (contact info below). Bottom line, even at genome-wide scale, they're very affordable.
1. What is a pooled-format CRISPR knockout screen?
Let's break this down.
By "screen" we mean a large-scale, unbiased study designed to isolate the subset of cells (or for in vivo screens, individuals) with mutations that result in a given phenotype of interest.
By "pooled-format" we are referring to the fact that the reagents used to make the mutations are introduced into cells in bulk and at random, rather than separately and systematically in something like a 96-well plate. In your 'pool' you'll have lots of different mutations represented. The key to pooled-format screening is figuing out how to separate the cells that have a phenotype of interest from everything else. We'll discuss how this is done later.
By "CRISPR knockout" we're referring to the fact that with the strategy we're talking about, the CRISPR-Cas system is used to make inactivating mutations in target genes. With the strategy we developed, this means targeting genes with specific sgRNAs. A "library" of different sgRNAs is introduced into cells that express Cas9. For the strategy we support, the sgRNAs integrate into the genome, giving us a molecular foothold with which to match cells and their phenotypes with specific sgRNAs and thereby, with specific genes.
2. What types of screen assays are feasible to do in pooled format?
What's key to pooled-format screens is to have some way to separate cells with a phenotype from those without, and then collect cells in one, the other, or both set, so that the sgRNAs in that set can be amplified, sequenced, and compared with the full set of sgRNAs you started out with.
Wait, huh? Not clear? Explaining with examples is probably easiest.
Essential gene screens -- a.k.a. 'Dropout' assays. Before you screen, you have your pool of cells, each of which has a different sgRNA and thus can have a mutation in a different gene. Let's say you want to identify essential genes; that is, the set of genes that when knocked out, result in cell death or growth arrest. To do this, you'd introduce your library, wait several weeks, and collect the cells that are left. Then, you'd compare what sgRNAs are present in the end population versus at the beginning. If an sgRNA targets an essential gene, then you'd expect that sgRNA to 'drop out' from the pool as you grow the cells over time. A screen like this was described in Viswanatha et al. (2018). References are linked below.
Selections -- a.k.a. toxin resistance assays. But okay, we're not all looking for essential genes. Let's say, differently, you're interested in how a particular cytotoxic drug or other toxin works. To what protein does it bind? How does it get into cells? And so on. To uncover that type of biology, you might choose to do a selection assay. With this workflow, you'd introduce the library of cells, treat the cells with your toxin (one or more times), and then let them grow for a while. If an sgRNA targets a gene that is required for the toxin to enter cells or otherwise be toxic to cells, then you'd expect that knockout of those genes would confer resistance, and thus, sgRNAs targeting these genes should be enriched in the toxin-treated population as compared to a mock-treated or untreated control pool of cells.
Everything else -- using FACS and other methods to sort this from that. We can't all be so lucky that our biological topic of interest is a good match for an essential gene screen or a toxin resistance assay. Not to worry! There are other options. Any way you can separate cells with your phenotype of interest from the rest would work. One way to do this would be to use FACS to sort for a marker or a reporter. We're developing assays like this and actively seeking collaborations. If you need help with this type of assay, or have other assay ideas you'd like to run by us, please feel free to reach out. Contact info is below.
Well, everything else, except ... If you're interested in an image-based readout, such as the sub-cellular localization of a specific biomolecule, then pooled-format screens are probably not the best fit, unless maybe you have access to a nifty new instrument that combines flow cytometry with imaging and image analysis (we don't but we hear that such marvels are out there somewhere!). For image-based readouts, the tried-and-true arrayed-format RNAi screens would be a fit. We support RNAi screens, too. Nose around on this site and you'll find protocols, information about past screens, available cell lines (including some GFP-tagged to mark specific organelles), and much more. To get you started, here's a link to our list of available arrayed-format Drosophila RNAi screen libraries: https://fgr.hms.harvard.edu/fly-cell-rnai-libraries
3. What is the typical screen workflow?
Some of the hard part of screening is already done! You can order 'screen-ready' cells from the DGRC and sgRNA libraries from Addgene.
Once you have both screen-ready cells (i.e., cells that express Cas9 and have attP sites) and one or more sgRNA library, the next step is to use transfection to introduce the library into cells. Then, you'd grow the cells the appropriate amount of time for your assay--typically, several weeks--and do any necessary treatments.
Along the way, and particularly at the start and finish of your assay, you'll collect cells that you can use to prep genomic DNA and then amplify sgRNAs by PCR. The PCR products are then prepped for next-generation sequencing (NGS) and sent off to a core facility.
Once the NGS data are back, you can use any of a few analysis pipelines to identify your sgRNA-level results and identify the subset of high-confidence hits at the gene level.
Not part of the workflow itself but critically important is to know what you'll do once you identify gene 'hits' (results) in the cell-based screens. Will you test the genes using an established in vivo Drosophila assay or in mammalian cells? Integrate the results with other datasets? You'll want to have some answers to the "what comes next" question before you start screening.
4. How long will it take from start to finish?
It depends. (Of course that's the answer!)
But okay, we can be a little more specific.
A lot of time went into development and optimization of the cells and approach, sgRNA library design and cloning, and other pre-screening tasks. If you're using our cells and libraries, that's done already. But you will need to spend some time optimizing the assay. If you're using a toxic compound, for example, you'll want to consider how much to add and on what schedule, and if you're using FACS, you'll want to be sure that sorting does a good job of separating the one or more populations you're interested to collect and analyze. Very roughly speaking, let's say 1-3 months for assay development.
Then there's the screen itself. Very roughly speaking again, this takes 1-2 months. Another week or more to prep sgRNAs for NGS, and then weeks or so, depending on core facility turn-around times, to get your data.
Altogether, a cell-based pooled screen will take a real investment of time. One thing to note when you think about the screen timeline is that a lot of this time is time spent waiting, or passaging cells now and then. It's not weeks and weeks of daily, active at-the-bench work. (Also, fair warning: the screen will seem to fly by. But waiting to have NGS data back so you can start your data analysis? That's going to feel like an eternity!)
5. What if I want to use a different Drosophila cell type? Or a different insect cell type?
There's every reason to think that other Drosophila cell types will work. You'd have to choose a cell line with attP sites already present -- the DGRC has made several -- or build your own, and, you'd likely have to do the work of introducing Cas9 into the cell line and confirming that it's functional. That's all stuff we can advise on.
We've developed the system for some mosquito cell lines. Some of these are available at the DGRC, too. If you want to develop it for a new cell line, you'd have to start by using our Minos-based method or another method to introduce attP sites into the cell genome, and you'd probably had to identify a good pol III promoter as well. To get a sense of what's needed for this, check out Viswanatha, Mameli at all (2021). References below.
6. What's been published on this?
This paper from Viswanatha et al. (2018) in eLife reports development of the technique, a genome-wide essential gene screen, and other good stuff: https://elifesciences.org/articles/36333
This paper from Viswanatha et al. (2019) in Current Protocols in Molecular Biology provides a step-by-step protocol: https://currentprotocols.onlinelibrary.wiley.com/doi/abs/10.1002/cpmb.111
This paper from Viswanatha, Mameli, et al. (2021) in Nature Communications reports develompent of the system for use in mosquito cell lines: https://www.nature.com/articles/s41467-021-27129-3
7. How can I get advice and other support?
Remember what we mentioned at the beginnning: For Drosophila cell screens, we've made the 'screen-ready' cells available at DGRC (cell line #268) and the sgRNAs libraries needed for this approach at Addgene (134582, 134583, 134584). Reagents are also available for mosquito cell screens. You can find some 'screen-ready' mosquito cell lines at the DGRC for example.
In addition, we're here to provide advice and other support along the way.
Please feel free to reach out to us at any stage in the process. The main contact for collaborations and general 'hey, can someone please give me some advice?' inquiries is DRSC-BTRR Director Dr. Stephanie E. Mohr. You can contact her (well, me! I'm the author of this page!) at stephanie underscore mohr at hms (for Harvard Med School) dot harvard (can you tell I don't like spam emails even though I welcome yours?) dot edu.