Generate the multiplex CRISPR screening library
1. Linearize the GPS expression vector:
4 µg vector DNA
3 µl CutSmart Buffer (NEB)
3 µl I-SceI (NEB)
Add ultrapure water to 30µl
Mix thoroughly and incubate at 37ºC for 3 hours.
2. Clean up the cut vector:
Add 0.8X Ampure XP beads (Beckman Coulter), and perform the clean-up as per the manufacturer's protocol. Elute from the beads in 20ul ultrapure water.
3. Amplify the sgRNA cassette by PCR:
Set up a PCR reaction to amplify the U6-sgRNA cassette from the sgRNA expression library:
10 µl 5X Q5 Buffer (NEB)
1 µl dNTPs (NEB)
10 ng Template DNA
1 µl Forward primer (25 µM, IDT)
1 µl Reverse primer (25 µM, IDT)
0.5 µl Q5 polymerase (NEB)
Ultrapure water to 50 µl
Thermocycle under standard conditions as recommended by the manufacturer for 30 cycles.
4. Purify the PCR product by gel extraction:
Run the PCR product on a 1.5%-2% agarose gel. Excise the band corresponding to the size of the amplified U6-sgRNA fragment. Extract the DNA using the QIAEX II gel extraction kit as per the manufacturer’s recommendations. Elute from the beads in ~10 µl ultrapure water.
5. Perform Gibson assembly reaction to insert the sgRNA expression cassette into the GPS vector:
Set up a Gibson assembly reaction:
1 µg cut vector DNA
50 ng purified PCR product (consult NEB recommendations depending on the size of the vector and insert)
Add 2X HiFi assembly mix
Mix thoroughly and incubate at 50ºC for 1 hour.
6. Clean up the assembly reaction:
Add 0.8X Ampure XP beads (Beckman Coulter), and perform the clean-up as per the manufacturer's protocol. Elute from the beads in 10 µl ultrapure water.
7. Transform at high efficiency:
Pre-chill electroporation cuvettes and one vial of DH10B electrocompetent cells on ice. Add 2 µl DNA to 25 µl bacteria in a pre-chilled eppendorf tube, mix gently by flicking the tube, and then pipette carefully into the bottom of a cuvette. Electroporate (1.8 kV, 200 Ω, 25 µF), recovering as quickly as possible in 975 µl S.O.C. medium. Transfer into a shaking incubator, ideally at 30ºC for 90 mins, or if not possible 37ºC for 1 hour.
Spread all of the bacteria across 2 large (15 cm) LB-agar plates supplemented with the appropriate antibiotic (typically ampicillin or carbenicillin). Incubate overnight at 30ºC. In order to estimate the transformation efficiency, a set of dilution plates are also required: for example, plate out 1/1000th, 1/10,000th and 1/100,000th of the bacteria on standard (10 cm) LB-agar plates and incubate overnight at 30ºC; the next morning, count the colonies present and multiply by the dilution factor to estimate the overall number of colonies. Ideally, this should be at least 100-fold greater than the number of substrate-sgRNA combination in the multiplex library.
8. Purify the plasmid DNA:
Add ~10 ml LB to the large bacterial plates. Using a cell lifter, gently scrape off all of the bacterial lawn and transfer to a 50 ml tube. Repeat the process once and transfer into the same tube. Pellet the bacteria by centrifugation (3000 x g, 10 mins); the dry pellet can be stored at -20ºC. Extract the plasmid DNA using a midiprep/maxiprep (depending on the mass of the bacterial pellet) kit.
9. Verify library construction through Sanger sequencing:
Submit the midiprep/maxiprep DNA for Sanger sequencing, using sequencing primer(s) that cover both ends of the insertion site. In addition, grow-up multiple (eg. 4-10) colonies from the dilution plates overnight at 37ºC in a shaking incubator. Purify the plasmid DNA by miniprep, and perform Sanger sequencing to verify that the individual colonies contain the U6-sgRNA cassette correctly inserted.
Package and titre lentivirus
10. Seed HEK-293T cells:
Expand stock of HEK-293T cells. Trypsinise, count, and seed ~8-10 million cells per 15 cm plate, aiming for a 70-90% confluent plate the next morning.
11. Transfect to package lentivirus:
Transfect the HEK-293T cells using PolyJet transfection reagent to package lentivirus. Follow the manufacturer’s protocol, using 20 µg DNA (made up of 10 µg multiplex CRISPR screening vector from step 9 and 10 µg of packaging vector mix) in 0.5 ml DMEM and 60 µl of PolyJet reagent in 0.5 ml DMEM per 15 cm plate.
12. Harvest lentivirus:
Replace the media on the transfected cells 16-24 hours later. Collect the viral supernatant 48 hours post-transfection, remove cell debris by passing it through a 0.45 µm filter, and store in single-use aliquots at -80ºC.
13. Transduce target cells to titre lentivirus:
The optimal approach here will depend on the cell type of the target cells and the size of the multiplex library. For screens performed in HEK-293T cells, we seeded 2 million cells in 3 x 10 cm plates, and then 24 hours later (at ~4 million cells per plate) transduced them by adding increasing amounts of virus (eg. 200 µl, 400 µl, 800 µl).
14. Assess transduction efficiency via flow cytometry:
Culture the cells for at least 48 hours to allow for maximal transgene expression. Then trypsinise the cells, transfer a small aliquot to a 5 ml FACS tube, and analyse the cells by flow cytometry. To ensure that the vast majority of cells have no more than a single lentiviral integration, the optimal amount of virus should be that which yields ~25% transduced cells.
Perform multiplex CRISPR screen
15. Large-scale transduction of Cas9-expressing target cells:
The optimal approach here will depend on the cell type of the target cells and the size of the multiplex library. For screens performed in HEK-293T cells, we seeded 8 million cells stably expressing Cas9 per 15 cm plate, and then 24 hours later (at ~16 million cells per plate) transduced them by adding the predetermined amount of virus expected to yield ~25% transduction (typically around 1 ml). The number of plates required will depend on the size of the multiplex library; ideally, the number of transduced cells should be at least 100-fold greater than the number of substrate-sgRNA combinations.
16. Puromycin selection to eliminate untransduced cells:
After ~48 hours, combine all of the transduced cells into a single vessel and mix thoroughly. Discard half of the cells. Take out a small aliquot into a FACS tube and analyze by flow cytometry to verify the desired transduction efficiency has been achieved. Return the other half to the incubator for continued culture in the presence of puromycin (we used a concentration of 1.5 µg/ml for HEK-293T) to remove untransduced cells. Ensure the cells are in the appropriate number of plates/appropriate volume of media for optimal growth in order to allow efficient puromycin selection.
Continue to culture the cells in the presence of puromycin until the day of the sort, ensuring that sufficient numbers to maintain at least 100-fold representation of the library are kept throughout.
17. FACS to isolate stabilised cells:
At ~7 days post-transduction, combine the transduced cells. Wash the cells once with PBS, and then resuspend the cells at ~10 million cells/ml in sort solution (PBS + 0.2% FBS) and filter through a cell strainer. Take to FACS facility to extract the top ~5% of GFP++ cells, but retain an aliquot of cells to form the unsorted reference sample. Ensure that sufficient cells are sorted/collected to maintain at least 100-fold representation of the library in each case.
Generate Illumina library
18. Extract genomic DNA:
Spin down the sorted cells (800 x g, 5 mins), resuspend the pellet in ~1 ml PBS and transfer to a 1.5 ml Eppendorf tube. Repeat the spin, and remove the PBS. Then follow the Puregene Cell Kit manufacturer’s protocol in order to extract the genomic DNA from the sorted cells. Repeat a similar process to extract the genomic DNA for the unsorted library cells. Analyze the resulting DNA by nanodrop spectrophotometry to quantify the DNA concentration.
19. Amplify substrate-sgRNA pairs by PCR:
Perform an initial PCR reaction to amplify the substrate-sgRNA pairs from the lentiviral construct. Set up the following on ice:
10 µl 5X Buffer (Agilent)
1 µl dNTPs (NEB)
2 µg Template DNA
1 µl Forward primer (25 µM, IDT)
1 µl Reverse primer (25 µM, IDT)
0.5 µl Herculase II polymerase (Agilent)
Ultrapure water to 50 µl
(Set up multiple reactions as required in order to amplify enough template genomic DNA to represent at least 100-fold coverage of the library). Thermocycle under standard conditions as recommended by the manufacturer for 23 cycles.
20. Clean up the PCR1 reaction:
Using the Qiagen PCR purification kit as per the manufacturer’s protocol, clean-up the PCR1 reactions. If multiple PCR1 reactions were performed, ensure that these are combined and mixed thoroughly first. Elute in 30 µl ultrapure water and quantify using a Nanodrop spectrophotometer.
21. Perform PCR2 reaction to generate Illumina-compatible library:
Perform a second PCR reaction to generate an Illumina compatible library, using 200 ng of the first PCR product as the template. PCR2 primers contain the Illumina P5 (and potentially i5) sequences on the forward primer, and Illumina P7 and i7 sequences on the reverse primer. Set up the following on ice:
10 µl 5X Buffer (Agilent)
1 µl dNTPs (NEB)
200 ng PCR1 product
1 µl P5 forward primer (25 µM, IDT)
1 µl P7 reverse primer (25 µM, IDT)
0.5 µl HerculaseII polymerase (Agilent)
Ultrapure water to 50 µl
Thermocycle under standard conditions as recommended by the manufacturer for 7 cycles.
22. Prepare Illumina library:
Clean up the PCR2 reactions exactly as in step 20 above. Elute in 30 µl ultrapure water and quantify using a Nanodrop spectrophotometer. On the basis of the Nanodrop readings, pool the samples in the desired ratio. Run the sample on a 1.5-2% agarose gel, and excise the band of the appropriate size. Extract the DNA using the QIAEX II Gel Extraction Kit, eluting from the beads in ~20 µl ultrapure water.
23. Perform Illumina sequencing:
Submit the library for Illumina sequencing. Paired-end reads are required in order to cover the substrate (forward read) and sgRNA (reverse read). The appropriate read depth will depend on the design of the experiment, but as a minimum would be equivalent to the number of cells sorted in step 17.
Analyse Illumina data
23. Map Illumina reads:
A sample Python script to enable the mapping of multiplex CRISPR screening data is provided1. This will convert the Illumina sequences reads stored in gzipped fastq files into raw count tables enumerating the frequency with which each substrate-sgRNA combination was observed in each sample. These can then be further analyzed using the MAGeCK algorithm to identify substrate-sgRNA combination enriched in the sorted cells.