CRISPRi/a Screening with Human iPSCs

Abstract

Identifying causative genes in a given phenotype or disease model is important for biological discovery and drug development. The recent development of the CRISPR/Cas9 system has enabled unbiased and large-scale genetic perturbation screens to identify causative genes by knocking out many genes in parallel and selecting cells with desired phenotype of interest. However, compared to cancer cell lines, primary human somatic cells including cardiomyocytes (CMs), neuron cells, endothelial cells are not easy targets of CRISPR screens because CRISPR screens require a large number of isogenic cells to be cultured and thus primary cells from patients are not ideal. The combination of CRISPR screens with induced pluripotent stem cell (iPSC) technology would be a powerful tool to identify causative genes and pathways because iPSCs can be expanded easily and differentiated to any cell type in principle. Here we describe a robust protocol for CRISPR screening using human iPSCs. Because each screening is different and needs to be customized depending on the cell types and phenotypes of interest, we show an example of CRISPR knockdown screening using CRISPRi system to identify essential genes to differentiate iPSCs to CMs.

1. Introduction

Although recent development in next-generation sequencing (NGS) has made it much faster and easier to study genome-wide gene expression changes and epigenetic changes even on a single-cell level, it’s not straightforward to study causality of a biological phenotype from the large datasets because they are mostly consequences of upstream changes [1–3]. Thus, identifying causative genes and pathways for biological discovery and drug development is still a time-consuming process. To accelerate identification of causative genes and pathways in biological phenotypes of interest, systematic and high-throughput genetic perturbation approaches have been done using chemical DNA mutagens or shRNA libraries. However, these methods had still limitations of efficiency and accuracy [1–3]. Recent CRISPR/Cas9-based systems enabled more efficient and accurate gene knockout screening even in a genome-wide scale [3–7] (Fig. 1). Transcriptional inhibition or activation screening is also available by using nuclease-deactivated Cas9 (dCas9) [1, 2, 8, 9]. These CRISPR/Cas9-based screening methods have been used mainly in cancer cells to identify essential genes for proliferation or resistance to anti-cancer drugs [1–3, 7, 9] (Fig. 2A).

Figure 1. CRISPR screening in combination with iPSC technology.

Identifying causative genes is important for biological discovery and drug development but has been time-consuming. On the other hand, forward genetic screening with pooled CRISPR libraries enables unbiased identification of causative genes even in genome-wide scale by knocking out (or knocking down or activating) many genes at once and selecting cells with a phenotype of interest. Thus, CRISPR screening in combination with iPSC technology, which can make any cell type in principle, is a powerful approach to identify causative genes of diverse phenotypes in human somatic cells.

Figure 2. Unbiased genetic screening using CRISPR sgRNA libraries.

(A) CRISPR screening can detect causative genes in a phenotype of interest by comparing the distributions of sgRNAs. After each cell is labelled by a sgRNA using pooled lentiviral library, the cells with a phenotype of interest are selected. Then, the distribution of sgRNAs in each population can be measured by next-generation sequencing. (B) We show an example schedule of CRISPRi screening with iPSC-CMs that identifies essential genes for CM differentiation. The CRISPRi iPSCs were infected with a sgRNA library and underwent puromycin selection. Then, iPSCs were differentiated to CMs. To screen essential genes for CM differentiation, CMs were stained by TNNT2 and sorted by flow cytometry. The volcano plot shows multiple hit genes and a positive control (TNNT2) as essential genes for CM differentiation.

Although forward genetic screening with loss-of-function or gain-of-function is a powerful approach to identify causative genes, it is not easy to apply CRISPR screening to human somatic cells because CRISPR screening requires a large number of cells to be cultured. Typically, the number of cells needs to be more than 1,000-fold of the number of gRNAs in a systematic, high-coverage screen to avoid the uncovered bias (statistically, small sample size leads to a high variability) [1]. For example, for genome-wide libraries which contain 100,000 gRNAs, assuming that the transfection efficiency of the gRNA library is 30%, more than 300 million cells need to be cultured. Thus, human primary cells are not ideal for large-scale screening. On the other hand, human iPSCs are a good platform to perform CRISPR screening because iPSCs are expandable and can be differentiated to any cell type in principle [10, 11] (Fig. 1). Here we describe an example protocol for CRISPR screening with human iPSCs. It is difficult to generalize the protocol of CRISPR screening with iPSCs because each screening is unique and different depending on the purposes, cell types of interest, phenotypes of interest, and types of genome editors (Cas9, dCas9, base editors, etc) [1]. In this chapter, we show an example protocol of CRISPRi screening to identify essential genes during CM (cardiomyocyte) differentiation. This protocol is modifiable to fit many other applications (Fig. 2).

2. Materials

2.1. CRISPRi/a library

1. Human Genome-wide CRISPRi-v2 Libraries or your custom library (Addgene, #83969) or Your custom sgRNA library: follow the protocols in Addgene #83969

2. Escherichia coli MegaX DH10B T1^R Electrocomp Cells (Thermo Fisher Scientific, C640003)

3. Escherichia coli Stellar Competent Cells (Clontech, 636763)

4. Escherichia coli DH5a competent cells (e.g., Zymo Research, T3007)

5. Large plate for library amplification (e.g., Fisher Scientific, 12-565-224)

6. LB Agar

7. LB medium

8. Ampicillin

9. Plasmid Midi prep kit

10. Plasmid Maxi or Giga prep kit

2.2. CRISPRi/a Plasmids

1. sgRNA backbone plasmid (pU6-sgRNA EF1Alpha-puro-T2A-BFP) (Addgene, #60955)

2. Lentiviral CRISPRi plasmid (UCOE-SFFV-dCas9-BFP-KRAB) (Addgene, #85969)

3. Lentiviral CRISPRa plasmid (pHRdSV40-dCas9-10xGCN4_v4-P2A-BFP) (Addgene, #60903)

4. Lentiviral CRISPRa plamid (pHRdSV40-scFv-GCN4-sfGFP-VP64-GB1-NLS) (Addgene #60904)

5. BstX1 (Thermo Fisher Scientific, FD1024)

6. Blp1 (New England Biolabs, R0585S)

2.3. Lentivirus production

1. psPAX2 plasmid (Addgene #12260)

2. pMD2.G plasmid (Addgene #12259)

3. HEK 293T cells (Clontech, Lenti-X^™ 293T Cell Line, 632180)

4. PEI MAX (Polysciences, Linear Polyethylenimine Hydrochloride MW 40K, 24765-1)

5. Polybrene solution: 10mg/mL polybrene in sterile water

6. ViralBoost (ALSTEM, VB100)

7. DMEM high glucose

8. Fetal Bovine Serum (FBS)

9. Penicillin-Streptomycin (10,000 units/mL of penicillin and 10,000 μg/mL of streptomycin)

10. 0.45-um PES syringe filter

11. Vacuum filtration system (e.g., Millipore, S2GVU02RE)

12. 10-cm cell culture dish

13. T225 cell culture flask

2.4. iPSC culture

1. CRISPRi iPSC line: If you want to make a new CRISPRi or CRISPRa line from iPSCs you have, follow the protocol provided on Addgene (Addgene, #83969)

2. DMEM/F-12

3. Essential 8 Medium (Thermo Fisher Scientific, A1517001)

4. RPMI 1640

5. RPMI 1640 no glucose

6. Matrigel Matrix Basement Membrane (Corning, 356231)

7. Y-27632 2HCl (ROCK Inhibitor) (Selleck Chemicals, S1049)

8. CHIR-99021 (Selleck Chemicals, S2924)

9. IWR-1 (Selleck Chemicals, S7086)

10. Puromycin (Thermo Fisher Scientific, A1113803)

11. EDTA

12. B-27 Supplement, minus insulin (Thermo Fisher Scientific, A1895601)

13. B-27 Supplement (50X), serum free (Thermo Fisher Scientific, 17504044)

2.5. Genomic DNA extraction

1. NK lysis buffer (50 mM Tris, 50 mM EDTA, 1% SDS, pH 8)

2. Proteinase K (Qiagen, 19131)

3. RNaseA (Qiagen, 19101)

4. Ammonium acetate (Sigma-Aldrich, A1542)

5. Isopropanol

6. Ethanol, molecular biology grade

7. 1x TE solution pH8.0

8. NanoDrop (Thermo Fisher Scientific)

2.6. NGS library prep and sequencing

1. KAPA HiFi HotStart DNA Polymerase with 5X Fidelity Buffer KK2502 (Fisher Scientific, NC0636151)

2. PCR Thermal Cyclers

3. QIAquick PCR Purification Kit (Qiagen, 28106)

4. 3M Sodium acetate

5. Agencourt AMPure XP (Beckman Coulter, A63881)

6. DNA LoBind tube (Fisher Scientific, 13-698-791)

7. Access to Bioanalyzer (High sensitivity DNA chip)

8. Access to Illumina NextSeq (or HiSeq)

2.7. Data analysis

1. ScreenProcessing (GitHub, https://github.com/mhorlbeck/ScreenProcessing)

2. Access to Python 2.7.

3. Methods

Here we show an example protocol for CRISPRi screening with human iPSCs. It can be modified to other types of screening including activation screening (see Note #1). Make sure to start with optimization in a small-scale experiment of 2–3 sgRNAs instead of directly proceeding to a screening that requires a large number of cells (see Note #2).

3.1. Prepare a CRISPRi iPSC line

3.1.1. Prepare a CRISPRi iPSC line and optimize culture condition

Prepare a CRISPRi iPSC line which stably expresses dCas9-KRAB. To generate a new CRISPRi iPSC line by lentivirus (transgenic) or genome editing (knock-in), follow the protocol provided by Weissman lab on Addgene [8, 9, 12]. Here we use an inducible dCas9-KRAB knock-in iPSC line generated by polyclonal lentivirus transduction [8]. Once you have a CRISPRi iPSC line, optimize culture condition and differentiation protocol. We use Essential 8 (E8) medium on Matrigel-coated (1:200 dilution) 6-well plates to maintain iPSCs. For passaging of iPSCs, we detach iPSCs from the plates with 0.5mM EDTA and culture them in E8 medium with 10 uM Y-27632 ROCK inhibitor (E8+Y medium) for 24 h. To differentiate CMs from this iPSC line, we use 8 uM CHIR-99021 in RPMI with B27 minus insulin supplement for 48 h and then 5 uM IWR-1.

3.1.2. Decide the concentration of puromycin for drug selection

It is important to find an appropriate concentration of puromycin using your iPSC lines. For screening, too low concentration may cause too many sgRNA-negative cells in your samples. Too high concentration may increase the percentage of cells that have 2 or more sgRNAs. If you plan to do drug selection after differentiation, you need to test the concentration in your cell type.

1. Day 0: Plate iPSCs in a Matrigel-coated 24-well plate (40,000 cells/well) in E8+Y medium. Culture cells overnight.

2. Day 1: Start puromycin. Try different concentration of puromycin (0.1 uM - 5 uM) in E8 medium (400 uL/well).

3. Continue puromycin selection for 3 days. Replace medium every day.

4. Day 4: Evaluate the confluency of cells. Find the lowest concentration that killed all the iPSCs.

3.2. Test small-scale transduction and lentiviral infection

Before starting a large-scale screening, confirm the knockdown efficiency in the cell type you will use in the screening afterwards because cell types affect chromatin accessibility of genome editors (e.g., dCas9-KRAB). Even if the knockdown efficiency is high in stem cell state, it does not guarantee a high efficiency in your cell type after differentiation. Small-scale experiments to evaluate the knockdown efficiency in your desired cell type targeting 2 or 3 genes are highly recommended.

3.2.1. Small-scale lentiviral production.

1. Prepare lentiviral plasmids that expresses sgRNAs (backbone: Addgene#60955). You need a non-targeting sgRNA in addition to sgRNAs that target your gene of interest (e.g., control sgRNA, sgRNA#1, and sgRNA#2). To design sgRNAs, use CRISPRiaDesign on GitHub or pick up the sequences from the Human Genome-wide CRISPRi-v2 Library (Addgene#83969) [12].

2. Culture HEK293T cells in DMEM 10% FBS (DMEM GlutaMax with 10% FBS and Penicillin/Streptomycin). On the day before transfection (Day 0), plate 800,000 cells per well in a 6-well plate (Table 1).

3. Day 1: For each sgRNA, prepare a lentiviral plasmid mixture and PEI MAX mixture (A and B below) and wait for 5 min. Mix A and B and incubate for 20 min in room temperature. Then gently drop the mixture to HEK293T cells in one well of a 6-well plate. Return the plate to the incubator and culture cells for overnight.

   	Components	for each sgRNA
   A	DMEM (serum free)	250 uL
   	psPAX2	1.8 ug
   	pMD2G	0.6 ug
   	lenti sgRNA plasmid	1.8 ug
   B	DMEM (serum free)	250 uL
   	PEI MAX	8.4 uL

4. Day 2: Check BFP (Blue fluorescent protein) on a fluorescent microscope to confirm PEI transfection is successful. Replace the medium with 2 mL of E8 medium per well. Culture cells for 48 h.

5. Day 4: Collect supernatant from each well. Filter the virus supernatant using 5-ml syringes with 0.45-um (or 0.22-um) syringe filters. Freeze the filtered supernatant in −80°C.(Optional) Using E8 medium on Day 2 as basal medium for virus production enables you to skip the virus concentration step and avoid exposing iPSCs to FBS. The functional titer of the supernatant virus is normally high enough without concentration. If you need higher titer and purity, you can concentrate virus with concentration reagents (e.g., Lenti-X Concentrator, Clontech).

Table 1.

Small-scale lentivirus production (6-well plate)

Day 0 Prepare HEK293T	Day 1 PEI transfection	Day 2 Check BFP	Day 4 Collect virus
• Plate HEK293T cells in a 6-well plate (800K cells/well for each sgRNA).	• Replace medium (1.5 mL, DMEM 10% FBS). • Prepare PEI mixture. • Incubate 20 min. • Gently drop into each well	• Check BFP. • Replace medium (2 mL, E8 medium). • Incubate for 48 h.	• Collect supernatants. • Filter supernatants. • Make aliquots and freeze them in −80°C.

3.2.2. Small-scale lentiviral infection of iPSCs.

1. Culture iPSCs on a Matrigel-coated 6-well plate until they reach 80% confluency and then proceed to lentiviral infection (Table 2).

2. Day 0 (iPSC plating and lentivirus infection): Detach iPSCs with 0.5mM EDTA and suspend them in E8+Y medium in the same way with regular iPSC passaging. Plate iPSCs to a new Matrigel-coated 6-well plate (300,000 cells/1.5mL/well = 30,000/cm², one well for each sgRNA). If you have 3 sgRNAs, prepare 3 wells (e.g., control sgRNA, sgRNA#1, and sgRNA#2). The cell density can be increased up to 600,000 cells/well depending on the growing speed of your iPSC line (see Note #3).

3. Add 1.6 uL of polybrene 10 mg/mL (final concentration 8 ug/mL).

4. Add 500 uL of lentivirus supernatant (total volume 2 mL). Return the 6-well plate to the incubator and wait for 3–4 h.

5. Remove the virus-containing medium and feed the cells with 2 mL of E8+Y medium. Culture cells overnight.

6. Day 1: Replace medium to E8 medium (2 mL).

7. Day 2 (Start puromycin selection): Check BFP on a fluorescent microscope (expect 40–50% positive) to confirm the lentivirus infection is successful. Then start puromycin at the concentration determined in 3.1.2 (typically, around 2 ug/mL).

8. Day 3: Replace medium (E8 with puromycin).

9. Day 4: Replace medium (E8 with puromycin).

10. Day 5: Finish puromycin selection. Typically, the cells reach around 80% confluency at this time point. Passage iPSCs to a new Matrigel-coated 6-well plate.

11. Day 6: Check BFP and expect 90% positive. If the BFP positive cells are still < 80%, start puromycin selection again for another 2 days.

12. Day7-: When the cells reach 80–90% confluency with 90% BFP positive, passage iPSCs to a new Matrigel-coated 6-well plate and also make frozen stocks.

Table 2.

Small-scale lentiviral infection of iPSCs (6-well plate)

Day 0 iPS plating & Lenti infection	Day 1	Day 2 Puromycin selection	Day 3
• Plate iPSCs (300K/well). • Add polybrene (final 8 ug/mL). • Add lentivirus 500 uL. • Incubate for 3–4 h. • Replace medium (E8+Y medium).	• Replace medium (E8).	• Check BFP (40–50% positive). • Start puromycin(2 ug/mL).	• Continue selection (E8+puromycin).
Day 4	Day 5 Passaging	Day 6	Day 7-
• Continue selection (E8+puromycin).	• Passage to a new plate.	• If BFP < 80%, start puromycin again.	• Make frozen stocks and start differentiation.

3.2.3. Evaluate the knockdown efficiency in your desired cell type.

1. Differentiate the sgRNA-expressing iPSCs to your desired cell type such as CMs or endothelial cells. Then, start doxycycline (2 uM) to express dCas9-KRAB if you are using Tet-On promoter.

2. Culture cells in as close to the screening condition as possible. Perform qPCR of the gene of interest after knockdown. Use a non-targeting sgRNA as control and multiple sgRNAs for the gene of interest. If knockdown efficiency is not high enough, you need to optimize conditions (the treatment period and timing of doxycycline, promoter of dCas9, promoter of sgRNAs, etc).

3.3. Large-scale lentiviral production for the pooled gRNA library

3.3.1. Prepare lentiviral sgRNA library.

1. Prepare CRISPRi sgRNA library (e.g., Addgene#83969). If you prefer to use custom sgRNA library to narrow down the target genes, follow the protocol provided from Weissman lab on Addgene (Addgene#60955, #83969) [12].

2. Amplify the sgRNA library. To keep the sgRNA distribution, plating on LB Agar plates is better than liquid amplification in LB medium.

3.3.2. Large-scale lentivirus production

1. Culture HEK293T cells in DMEM 10% FBS until they reach the enough number of cells to proceed. On the day before transfection (Day 0), plate 18 million cells per T225 flask and prepare at least 4 flasks. Culture cells overnight in DMEM 10% FBS (Table 3).

2. Day 1: Mix the lentiviral plasmids and PEIMAX as below and incubate for 20 min in room temperature.

   	per flask
   DMEM (serum free)	651	uL
   psPAX2 plasmid	6.8	ug
   pMD2G plasmid	3.4	ug
   Lenti library plasmid	13.6	ug
   PEIMAX	195	uL

3. Mix DMEM 10% FBS (25 mL per flask) and the transfection mixture. Gently feed the cells in each flask.

   	per flask
   transfection mixture	(above)
   DMEM 10% FBS	25	mL

4. Day 2: Check BFP on a fluorescent microscope to confirm the PEI transfection is successful.

5. Replace medium to E8 (25 mL) and incubate cells for 48 h. Make sure to keep the flasks steady and level to avoid drying in the incubator because the amount of medium is relatively small.

   (Optional) Adding ViralBoost (1:500 ratio) at this time point can increase the viral titer.

6. Day4: Collect supernatant from all the flasks and mix well. Filter the virus supernatant with 0.45-um (or 0.22-um) filter bottle. Then, make aliquots (10–25mL) and freeze them in −80°C (make one 3-mL aliquot for functional titration).

Table 3.

Large-scale lentivirus production

Day 0 Prepare HEK293T	Day 1 PEI transfection	Day 2 Check BFP	Day 4 Collect virus
• Plate HEK293T cells in T225 flasks (18M cells/flask).	• Prepare PEI mixture. • Incubate 20 min. • Feed the cells with DMEM 10% FBS containing PEI mixture.	• Check BFP. • Replace medium (25 mL, E8 medium).Optional: ViralBoost • Incubate for 48 h.	• Collect supernatant. • Filter supernatant. • Make aliquots and freeze them in −80°C.

3.3.2. Functional titration of lentiviral library

Perform functional titration of lentiviral library in as close to the screening condition as possible. You can use the same schedule with Day 0–2 in small-scale testing (Table 2).

1. Day 0 (lentiviral infection): Thaw one aliquot of lentiviral library (3 mL is enough).

2. Plate CRISPRi iPSCs on a 6-well plate (300K/well) in 1.5-mL of E8+Y medium. Add 1.6 uL of polybrene (final 8 ug/mL) in the same way with small-scale testing.

3. Make serial dilution by adding lentiviral library (0 uL, 31.25 uL, 62.5 uL, 125 uL, 250 uL, 500 uL) to each well. Supplement E8+Y medium to make the total volume 2 mL/well. Return the plate to the incubator and wait for 3–4 h.

   Lentiviral library	0	31.25	62.5	125	250	500	(uL)
   E8+Y medium	500	468.75	437.5	375	250	0	(uL)

4. Remove virus-containing medium and feed the cells with E8+Y medium (2 mL). Culture cells overnight.

5. Day 1: Replace medium to E8 medium.

6. Day2 (flow cytometry): Check BFP on a fluorescent microscope to confirm lentiviral infection is successful. Detach the cells with 0.5mM EDTA and perform flow cytometry analysis to see the positive ratio of BFP. The ideal functional titer is 30%. Determine the appropriate concentration of lentiviral library. For example, if 62.5-uL virus in 2 mL medium (1:32 ratio) makes 30% BFP positive cells, 781.25 uL of lentivirus in 25 mL medium is good for 15-cm dish scale.

3.4. Large-scale lentiviral infection of iPSCs

3.4.1. Prepare the large-scale experiments

1. Calculate how many cells you need. The required cell number (sgRNA positive) is 1000 x of library size. For example, for a library with 10,000 sgRNAs, you need 33 million iPSCs (1,000 × 10,000 / 0.3) to be infected. For a genome-wide library with 100,000 sgRNAs, you need 333 million iPSCs to be infected (see Note #2).

2. Ensure you have enough E8 medium, 15-cm plates, and other reagents (Matrigel, ROCK inhibitor, etc) before starting cell culture. Keep incubator space for many plates.

3.4.2. Lentiviral library infection of iPSCs

1. Culture iPSCs on Matrigel-coated 15-cm dishes until they reach enough number of cells and proceed to large-scale lentiviral infection (Table 4).

2. Day 0 (iPSC plating and lentivirus infection): Detach iPSCs with 0.5mM EDTA and suspend them in E8+Y medium in the same way with small-scale test. Plate iPSCs to Matrigel-coated 15-cm dishes (4.5M/20 mL = 30,000/cm²). Make sure to prepare enough number of cells and plates to keep the coverage of 1000x as calculated above. (You can use a higher cell density, if you have optimized it in the small-scale infection, so that you can reduce the total number of plates. See Note #2 and #3.)

3. Mix polybrene, lentiviral library, and E8 medium as below. Gently drop 5 mL of virus mixture into each dish. For virus library, use the dilution ratio determined in functional titration (3.3.2). The total volume in each 15-cm dish will be 25 mL. Return all the plates to the incubator and wait for 3–4 h.

   	per a 15-cm dish
   CRISPR iPSCs (in E8+Y medium)	20	mL
   polybrene	20	uL (final 8 ug/mL)
   lentiviral library	x	uL
   E8+Y medium	5000-x	uL

4. Remove the medium containing lentiviral library and feed the cells with 25 mL of E8+Y medium. Culture cells overnight.

5. Day 1: Replace medium to E8 medium (25 mL).

6. Day 2 (Start puromycin selection): Check BFP on a fluorescent microscope to confirm the BFP positive ratio is about 30%. Then start puromycin at the concentration determined in 3.1.2 (typically, around 2 ug/mL).

7. Day 3: Replace medium (E8 medium with puromycin).

8. Day 4: Replace medium (E8 medium with puromycin).

9. Day 5: Finish puromycin selection. Typically, the cells reach 80% confluency at this time point. Detach iPSCs from all the plates with 0.5mM EDTA and suspend cells in E8+Y medium. Mix well so that you have a homogenous cell suspension here.

10. Count the number of cells with a cell counter (typically, 40–80 million per plate).

11. Plate iPSCs in new plates. To avoid uncovered bias, you need 1000x coverage of cell number. For example, if your library contains 10,000 sgRNAs, you need to plate more than 12.5 million (= 1000 × 10K / 0.8) cells, assuming you have 80% BFP positive cells.

    (Optional) You can make frozen stocks here. Again, you need 1000x coverage for each stock.

12. Day 6: Check BFP and expect 90% positive. If the BFP positive cells are only < 80%, start puromycin selection again for another 2 days.

13. Day7-: When the cells reach 80–90% confluency with 90% BFP positive, passage iPSCs to new plates and also make frozen stocks. Make sure to keep 1000x coverage for passaging and freezing.

Table 4.

Large-scale lentiviral infection of iPSCs

Day 0 iPS plating & Lenti infection	Day 1	Day 2 Puromycin selection	Day 3
• Plate iPSCs (4.5M/15-cm dish). • Make mixture of polybrene and lentiviral library. • Gently drop 5 mL of the virus mixture into each 15-cm dish. • Incubate 3–4 h. • Replace medium. (E8+Y medium)	• Replace medium (E8).	• Check BFP (30% positive). • Start puromycin(2 ug/mL).	• Continue selection (E8+puromycin).
Day 4	Day 5 Passaging	Day 6	Day 7
• Continue selection (E8+puromycin).	• Passage to new plates.	• If BFP < 80%, start puromycin again.	• Make frozen stocks and start differentiation.

3.5. Large-scale CRISPR screening

Perform screening using iPSCs that are infected with lentiviral library. Differentiate iPSCs to your desired cell type [10, 11, 13]. Start doxycycline, if you are using Tet-On promoter, at an appropriate time point depending on your purpose (It’s important to use appropriate controls because some cell lines may have leaky activity of Tet-On promoter). Select cells with your desired phenotype. The phenotype needs to be a screenable phenotype (cell proliferation, drug selection, flow cytometry, etc). If you need to stain intracellular proteins for flow cytometry, you can fix the cells with 4% PFA or BD Fixation/Permeabilization solution. You can keep the sorted cells in −20°C. Make sure to keep 1000x coverage through the experiment. We show an example schedule of CRISPRi screening to identify essential genes for CM differentiation (Fig. 2).

3.6. Genomic DNA extraction

Once you collect cells after drug selection or sorting, extract genomic DNA so that you can run PCR of sgRNA-coding sequences that are integrated to the genome of iPSCs by lentivirus. We recommend conventional isopropanol precipitation for genomic DNA extraction rather than spin column-based extraction to collect all the genomic DNA without omission [7]. The protocol below is for a sample of 30–50 million cells. If you have more cells per sample, scale up the amount of each reagent.

1. Make a cell pellet by centrifugation and add 4 mL of NK Lysis buffer in a 15-mL tube.

2. Add 20 uL Proteinase K (20 mg/mL) and mix well by inverting the tube more than 10 times. Incubate the tube at 37°C overnight.

3. Add 20 uL RNaseA (10 mg/mL) and mix well by inverting the tube more than 10 times. Incubate the tube at 37°C for 30 min.

4. Cool the sample on ice.

5. To precipitate proteins, add 2 mL of pre-chilled 7.5M ammonium and voltex the tube for 20 sec.

6. Centrifuge the tube at 3000 g for 20 min.

7. Genomic DNA is in the supernatant. Gently transfer the supernatant to a new tube.

8. Add 6 mL of isopropanol and mix well by inverting the tube more than 10 times.

9. Centrifuge the tube at 3000 g for 20 min.

10. Discard supernatant and add 6 mL of fresh 70%EtOH and mix well by inverting the tube more than 10 times.

11. Centrifuge the tube at 3000 g for 3 min.

12. Remove supernatant and dry the pellet for 10–30 min until the DNA pellet becomes slightly translucent.

13. Resuspend the pellet in 500 uL of 1x TE buffer (or 100 uL per 10 million cells).

14. To dilute the DNA pellet completely, incubate the sample at 65°C for 1 h in a heat block and then at room temperature (or 4°C) overnight.

15. Measure the concentration of DNA by NanoDrop (expect to obtain 6 pg of DNA per cell).

3.7. Next-generation sequencing (NGS) to determine sgRNA distribution.

To determine sgRNA distribution in your sample, you are going to run PCR for the sgRNA-coding sequences integrated to the genome and then quantify each sgRNA by NGS. In terms of sequencing platform, we use Illumina NextSeq in this protocol, but the structure of the NGS library should be compatible with other Illumina sequencers such as MiSeq and HiSeq. We recommend discussing the library structure and indexes with your core facility or sequencing service provider before running PCR because you cannot change the library structure and indexes after PCR. For sequencing depth, you need coverage of 500x. For a library size of 10,000 sgRNAs, you typically need 5 million reads per sample. For a genome-wide library of 100,000 sgRNAs, you need 50 million reads per sample. You can combine multiple samples in one NextSeq run by indexing. Typically, the number of reads from one NextSeq run is 300–400 million and you can submit at least 6–8 samples together by multi-indexing (see Note #4 for sequencing of plasmid library).

3.7.1. PCR of genomic DNA.

1. Clean your bench and keep a safe space for setting up PCR mixture. Tiny amount of contamination from sgRNA plasmid or library can spoil all the process you have done because your sample contains only about 1,000 copies per sgRNA. Don’t share the same tip box with plasmid works.

2. Set up PCR reactions below. Use one of the indexed 5’ primers and the common 3’ primer for each sample (Table 5 and 6). Scale up the number of reactions to amplify all the genomic DNA harvested from the screening samples. You can put up to 2.5 ug of genomic DNA in one PCR reaction. The forward primers contain TruSeq indexes to demultiplex samples after sequencing. You can use other TruSeq indexes to increase the number of samples (follow the protocol provided by Weissman lab on Addgene) [12].

   	per one reaction
   5X KAPA HiFi Fidelity Buffer	20	uL
   5’ Forward primer (100 uM)	1	uL
   Common 3’ primer (100 uM)	1	uL
   10 mM dNTP Mix	4	uL
   KAPA HiFi HotStart DNA Polymerase	2	uL
   PCR-grade water	up to 72	uL
   genomic DNA (up to 2.5 ug)	as required
   (total)	100	uL

3. Run PCR.

   Cycles	Temp.	Duration
   1	95°C	5	min
   23	98°C	20	sec
   	64°C	15	sec
   	72°C	15	sec
   1	72°C	10	min
   	4°C	inf

Table 5.

Forward PCR primers with TruSeq indexes

Sample #	TruSeq ID	TruSeq index	Primer sequence
1	12	CTTGTA	aatgatacggcgaccaccgagatctacacgatcggaagagcacacgtctgaactccagtcacCTTGTAgcacaaaaggaaactcaccct
2	6	GCCAAT	aatgatacggcgaccaccgagatctacacgatcggaagagcacacgtctgaactccagtcacGCCAATgcacaaaaggaaactcaccct
3	14	AGTTCC	aatgatacggcgaccaccgagatctacacgatcggaagagcacacgtctgaactccagtcacAGTTCCgcacaaaaggaaactcaccct
4	10	TAGCTT	aatgatacggcgaccaccgagatctacacgatcggaagagcacacgtctgaactccagtcacTAGCTTgcacaaaaggaaactcaccct
5	3	TTAGGC	aatgatacggcgaccaccgagatctacacgatcggaagagcacacgtctgaactccagtcacTTAGGCgcacaaaaggaaactcaccct
6	1	ATCACG	aatgatacggcgaccaccgagatctacacgatcggaagagcacacgtctgaactccagtcacATCACGgcacaaaaggaaactcaccct

Table 6.

Reverse PCR primer and NGS sequencing primer

	Primer sequence
Common 3’ PCR primer	CAAGCAGAAGACGGCATACGAGATCGACTCGGTGCCACTTTTTC
NGS sequencing primer	GTGTGTTTTGAGACTATAAGTATCCCTTGGAGAACCACCTTGTTG

3.7.2. Purification of amplified NGS library by QIAquick column.

1. After PCR, pool all your PCR reactions for each sample and mix well. All the sgRNAs in the genomic DNA should have been amplified about 8 million-fold (= 2²³) in principle. Thus, you can use only a fraction of the pool for purification and sequencing.

2. Transfer 200 uL of the pooled PCR reaction to a new 1.5-mL tube. You can keep the remaining as spare at −20°C.

3. Add 5 volume of Buffer PB (1,000 uL) in QIAquick PCR Purification Kit and 1/100 volume of 3M Sodium acetate (2 uL). Mix well.

4. Transfer the mixture to a QIAquick column. Centrifuge at 13,000g for 30sec. Repeat until all the sample goes through the column.

5. Discard flow-through. To wash, add 750 uL of Buffer PE to the QIAquick column. Centrifuge at 13000g for 30 sec.

6. Discard flow-through. Centrifuge again at 13,000g for 1 min to remove residual ethanol.

7. To elute DNA, add 125 uL of 1x TE buffer to the column, incubate at room temperature for 1 min. Set the column on a new 1.5-mL tube and centrifuge 13,000g for 1 min.

3.7.3. Size selection of NGS library by AMPure XP beads.

The PCR product is 275 bp. Perform a double size selection using AMPure XP beads. You may also use SPRIselect or Seramag speed beads (follow the protocol provided by Weissman lab on Addgene) [12].

1. Transfer 120 uL of purified sample to a new 1.5-mL DNA LoBind tube. Add 0.65 volume of AMPure XP beads (78 uL). Mix well and spin the tube briefly. Incubate at room temperature for 10 min.

2. Place the tube on a magnetic stand for 5 min until the supernatant becomes clear.

3. Transfer the supernatant to a new tube. Keep the supernatant.

4. Add 1 volume of AMPure XP beads (120 uL). Mix well and spin the tube briefly. Incubate at room temperature for 10 min.

5. Place the tube on a magnetic stand for 5 min until the supernatant becomes clear. Make sure to keep the tube on the stand until the final DNA elution step.

6. Remove the supernatant and keep the beads.

7. Wash the beads with 500 uL of freshly-made 80% EtOH. Incubate for 2 min and remove EtOH.

8. Repeat washing with 80% EtOH. Remove EtOH.

9. Incubate at room temperature for 10 min until the beads become dry.

10. Elute DNA in 22.5 uL of 1x TE buffer. Place the tube on the magnetic stand again to remove beads and transfer 20 uL to a new DNA LoBind tube.

3.7.4. Library QC and submission for sequencing.

1. Measure the yield with Bioanalyzer (High sensitivity DNA chip). Expect a clear peak around 275 bp. If you see background (broad low smear) around the peak, discuss it with your core facility or sequencing service provider. In many cases, the low smear does not affect sequencing results.

2. Based on the result of Bioanalyzer, calculate the concentration of each sample. Dilute the library and mix samples with different indexes if you are submitting multiple samples in one run. We usually prepare 15 uL of 5 nM for one NextSeq run.

   Sequencer	Illumina NextSeq
   Run type	Single-end 75 bp (SE 75)
   Index length	6 bp
   Read 1	custom primer (Table 6)
   Read 2	standard Illumina index primer
   PhiX spike-in	10%

3.7.5. Analyze the sequencing results.

1. While waiting for the sequencing results, install ScreenProcessing (GitHub, https://github.com/mhorlbeck/ScreenProcessing) [12] on Anaconda with Python 2.7 and test if the demo data provided in GitHub works in your environment.

2. Once you receive the fastq files, run fastqgz_to_counts.py to obtain count data of sgRNAs and follow the instruction provided in GitHub.

3. To obtain the statistical results of ScreenProcessing, run process_experiments.py [12].

3.7.6. After screening.

To obtain robust results, repeat 3–4 batches from viral infection part (see Note #5). After screening, perform validation for each gene that you are interested in. For validation of each gene, pick up multiple sgRNAs per one gene from your library and make sgRNA plasmids. You can use the same protocol with small-scale experiment (Table 1 and 2) for validation experiments.