For those living in the more Northern climes, this is just about the time when you would give nearly anything for a sunny pool party. When it comes to pool parties, CRISPR turns out to be a pretty good host, but with one important caveat – CRISPR pool parties tend to be life-or-death affairs.
Pooled CRISPR Arrays – Sink Or Swim
Pooled CRISPR array screening is a great way to find important genes in a pathway. You simply invite a bunch of cells by transforming them with a library of gRNAs, throw them together in a pool, apply a selective pressure, and see who keeps swimming. Cells will sink out of the pool if their inviting gRNA weighs them down by targeting a gene important in the selective condition. On the other hand, cells that receive innocuous gRNAs will be able to stay afloat. If you sequence all remaining cells (with their respective gRNAs) after selection, you will find that gRNAs targeting important genes are depleted from the pool.
The problem with this approach is that it works best for pathways that affect cell survival or – that other reason for pool parties – proliferation. You can also screen cells by fluorescence if you can fuse GFP to a particular marker gene, but that still leaves CRISPR pools fairly limited. To throw a bigger party, Paul Datlinger and Cristoph Bock in Vienna have developed CRISPR droplet sequencing, or CROP-seq.
CROP-Seq – CRISPR Screens For The Masses
CROP-seq improves on standard pooled CRISPR screening in the sequencing step. Instead of sequencing all the post-selection cells in bulk, the authors sequenced RNA transcription profiles from single cells via droplet sequencing (drop-seq). As proof of principle, they used CROP-seq to identify genes in the T Cell Receptor (TCR) activation pathway. After transforming Jurkat cells expressing Cas9 with a gRNA vector, they activated half the cells with anti-CD3 and anti-CD28 antibodies, and then individually droplet-sequenced each cell transcriptome.
After some data-cleaning, the team identified a TCR activation signature with a neat trick using principle component analysis (PCA) of the RNA profiles. The greatest source of variation was between naive cells and stimulated cells, so PCA identified that induction signature as the 1st principal component (PC1). Next, they aggregated transcriptomes by targeted gene, identified target genes with PC1 loading above the 99th percentile, and declared them “signature genes”. After validating with targeted bulk RNA sequencing and flow cytometry, the signature genes held up.
The advantage of CROP-seq is that gRNAs can be associated with single-cell transcriptome data, so you can sequence a huge library all at once. Two similar CRISPR pool parties recently popped up, Perturb-seq and CRISP-seq, but those both use transcribed barcodes to identify gRNAs. In comparison to those, CROP-seq is a bit easier on the pool party host, since it doesn’t require unique barcodes for each gRNA.
If winter weather is leaving you looking for a bit less crisp and a bit more pool party, jump on in at Nature Methods, Jan 2017.