Generating Mutant Renal Cell Lines Using CRISPR Technologies.

Gene editing using the CRISPR/Cas9 system is an extremely efficient approach for generating mutations within the genomic DNA of immortalized cell lines. This procedure begins with a straightforward cloning step to generate a single plasmid encoding the Cas9 enzyme as well as a synthetic guide RNA (sgRNA) which is selected to target specific sites within the genome. This plasmid is transfected into cells either alone, in order to generate random insertion-deletion alleles ("indels") at the desired locus via the nonhomologous end-joining pathway, or in conjunction with a homology-directed repair template oligonucleotide to generate a specific point mutation. Here we describe a procedure to perform gene editing in IMCD3 and HEK293 cells and to subsequently isolate clonal cell lines carrying mutations of interest.


Introduction
In recent years, there has been huge growth in the application of gene-editing technologies. The concept of gene editing began in the late 1990s with the development of zinc-finger nucleases (ZFNs) [1] and subsequently with transcription-like effector nucleases (TALENs) [2][3][4]. However, both of these technologies are limited by their relatively low targeting efficiencies, the relative rarity of suitable ZFN cut sites within the genome, and the need for complex cloning procedures to generate TALEN DNA-binding oligonucleotides. These limitations were circumvented by the development of the clustered regularly interspaced short palindromic repeats/CRISPR-associated endonuclease (CRISPR/ Cas9) system [5,6], following the observation that the type II CRISPR system from Streptococcus pyogenes (SpCas) could be engineered to edit mammalian genomes [7,8]. This editing system combines the Cas9 endonuclease with a synthetic guide RNA (sgRNA) that acts as a signal to deliver Cas9 to a specific genomic target sequence where it can induce a double-strand break (DSB; Fig. 1). Cas9 recognizes a protospacer adjacent motif (PAM) located immediately downstream of its sgRNA in order to elicit its function. DSBs can be repaired by one of the two mechanisms which act according to the presence or absence of a donor DNA template: nonhomologous end-joining (NHEJ) or homologydirected repair (HDR). NHEJ is error-prone, often introducing base insertions or deletions (indels), typically leading to a frameshift mutation. On the other hand, HDR employs a repair template with homology to the target site [9,10]. To verify the efficiency of Cas9 The sgRNA consists of a short region of homology coupled to a PAM sequence as well as a more complex region with secondary structure. The region of homology is all that is cloned into plasmid pX330 in this protocol. Cas9 cuts the DNA at the PAM sequence causing a double-strand break. This can be repaired either by the NHEJ pathway or, in conjunction with a template oligo, by HDR. These pathways typically result in the introduction of indels (red) or point mutations (blue) cleavage at the target site we employ the T7 endonuclease (T7EI) assay. T7EI recognizes and cleaves DNA heteroduplexes containing mutant DNA sequences following PCR amplification of the targeted site and reannealing, and serves as a simple means of gauging targeting efficiency and screening for clonal cell lines that carry mutations [11][12][13].
Here we describe a procedure which we have adapted from the Zhang Lab at the Broad Institute [9], in order to generate frameshift and point mutations in renal cell lines. Key features of this protocol include the following: (1) sorting of cells co-transfected with GFP together with the Cas9/sgRNA-expressing plasmid enhances the overall targeting efficiency. (2) We culture cells for at least 2 weeks following transfection of these plasmids before sorting single cells into clonal cell cultures, as we have found that shorter periods of time post-transfection are associated with a higher level of mosaic mutations, presumably owing to continued expression of Cas9/sgRNA. (3) For introduction of point mutations via HDR, we have found that a template oligo consisting of 60 nucleotides on either side of the DNA mismatch has good targeting efficiency (see Note 1).
To date, our lab has successfully employed these methods to produce immortalized renal cells carrying desired mutations (and defined genotypes) at over 60 genomic loci. We have predominantly used inner medullary collecting duct (IMDC3) cells, which are mouse epithelial cells allowing us to study the genes involved in cilia function, and human embryonic kidney (HEK293) cells, which have enabled us to model features of human renal diseases including polycystic kidney disease. We have also used this protocol successfully in other cell lines, including Chinese hamster ovary (CHO) cells, and the techniques could be used to edit genes in glomerular podocytes and endothelium, cells particularly affected in diabetic nephropathy. In other works, we have found that analysis of multiple clonal IMCD3 or HEK293 cell lines carrying biallelic mutations in genes of interest can have the advantage of reducing phenotypic variability, as compared to results generated using primary cell lines or transient gene-knockdown approaches, thereby revealing novel cellular phenotypes [14,15]. The improved phenotypic consistency of these cellular models means that the CRISPR/Cas9 system is invaluable for studies of disease mechanisms and for drug screening approaches.

Methods
The methods presented below should be used to generate random indels by NHEJ. A variation of this protocol can be used to generate a defined single base change mutation (or if required, substitution, insertion, or deletion of a few nucleotides) by HDR (see Note 1). For simplicity, the following methods describe the design and use of one sgRNA (see Note 2; Fig. 2 2. Select a 23-500-nucleotide (nt) genomic region that you wish to edit (see Note 3). Fig. 2 Overview of methodology. A sgRNA is designed and ordered as two oligos (1), which are annealed (2) and cloned into plasmid pX330 to yield plasmid pX330-sgRNA (3), which contains a Cas9 expression cassette and a sgRNA expression cassette. IMCD3 or HEK293 cells are transfected with pX330-sgRNA and pAcGFP1-C1, a GFP expression vector used as a positive marker of transfection (4). Fluorescence-activated cell sorting (FACS) is used for enrichment of GFP-positive cells in a step we term the "group sort" (5); these cells are presumed to have also successfully been transfected with pX330-sgRNA. Half of the group-sorted cells are reseeded and subcultured for 2 weeks (see Note 15). gDNA is extracted from the remaining cells and analyzed by T7 endonuclease I (T7EI) assay (6), to detect indels at the genomic target site within the cell population. After 2 weeks of subculture, the reseeded CRISPR cells are subjected to FACS again, and a single cell is seeded into each well of a 96-well plate, in a step that we call the "single-cell sort" (7) which is followed by clonal expansion (8). Each single cell has the propensity to establish a clonal colony, if it proliferates. Colonies are genotyped by PCR and Sanger sequencing to confirm the presence and nature of mutations in each (9) 3. Open the CRISPOR design tool (crispor.tefor.net). Select the appropriate target genome and paste your sequence into the search box. Submit the search. The online tool will produce a ranked list of all possible sgRNAs within the genomic sequence (see Note 4).
4. Identify a suitable sgRNA sequence from the displayed list (see Note 5).
5. Take the 20 nt sgRNA sequence (without the PAM). If its first nucleotide is not a "G," then add a "G" at the 5 0 end. This sequence will form the foundations of oligo 1.
6. Find the reverse complement of the sgRNA sequence (including the extra "G" nucleotide, if added). This will form the foundations of oligo 2.
7. To the 5 0 end of oligo 1, insert nucleotides "CACC." Design of oligo 1 is now complete.
3. Anneal in a thermocycler using the following cycling conditions: 95 C for 5 min, ramp down to 25 C at 5 C/min.
2. Incubate for 2 h at 37 C. 9. Verify the sequence of each plasmid by Sanger sequencing using the pX330 sequencing primer. We will call this plasmid pX330-sgRNA.

Transfection of HEK293 or IMCD3 Cells
A different transfection reagent is used depending on the cell line (see Note 7). Choose Subheading 3.3.2 for transfection of HEK293 cells and Subheading 3.3.3 for transfection of IMCD3 cells. Regardless of transfection reagent, three transfection conditions will be used: (1) a co-transfection with pX330-sgRNA and pAcGFP1-C1 ("CRISPR" transfection); (2) a transfection with pAcGFP1-C1 ("GFP-only" transfection); (3) a mock transfection without plasmid DNA ("non-transfected" condition). The latter two conditions act as positive and negative controls of transfection, respectively. 1. Label a 15 mL centrifuge tube for use as a CRISPR sample collection tube. Add 500 μL fresh medium to the tube, put it on ice, and set aside.
2. Wash the CRISPR and non-transfected wells of cells twice each with 1 mL/well sterile DPBS (see Note 10).
3. Add 1 mL/well trypsin-EDTA and incubate at 37 C until cells dissociate.
4. Add 2 mL/well complete media and transfer each 3 mL cell suspension to a 15 mL centrifuge tube.  8. Put the tubes directly on ice.
9. Proceed directly to FACS (see Note 13). Use non-transfected cells to discern single cells from cell clumps when setting up the cell sorter. Isolate GFP-positive cells from the CRISPR sample and collect these in the collection tube (prepared in Subheading 3.4, step 1). Run the entirety of the sample or collect at least 5 Â 10 5 cells per sample (see Note 14). Keep collected cells on ice.
10. In a sterile safety cabinet, take one-half of the GFP-positive CRISPR sample and seed them in a 25 cm 2 tissue culturetreated flask containing 5 mL complete medium. Return the cells to the incubator and continue to grow them for 2 weeks, passaging as and when necessary (see Note 15).
11. Separately pellet the remaining GFP-positive CRISPR cells and the non-transfected cells for gDNA extraction, by centrifugation at 300 Â g for 5 min. Discard the supernatant. Proceed directly to the next step, or store the cell pellets at À20 C.

Expansion of Cell Colonies
1. Incubate the cells for 2 weeks, to allow colonies to establish (see Note 24).

After 2 weeks, check individual wells under the microscope to
identify those that contain a cell colony. Monitor cell confluence in these wells over the ensuing days. 7. With the remaining flask, wash cells twice with 5 mL DPBS, and then dissociate in 1 mL trypsin-EDTA. Add 4 mL complete media and transfer the 5 mL cell suspension to a 15 mL centrifuge tube. Pellet the cells for 5 min at 300 Â g and discard the supernatant. Proceed directly to the next step, or store the cell pellet at À20 C.

Genotyping of Clones
This method describes the genotyping of a single-cell colony.
1. Extract genomic (g) DNA with the DNeasy Blood and Tissue kit, following the manufacturer's instructions.
2. Use the extracted gDNA as template to perform a 50 μL PCR reaction with the genotyping primers (designed in Subheading 3.1, step 10).
3. Run the 50 μL PCR in an agarose gel against a 1 kb DNA ladder. If there are multiple amplicons present, ensure that they are well resolved (see Notes 27 and 28).
5. Sanger sequence the extracted amplicon(s) to identify the presence and nature of mutations generated by CRISPR (see Note 29; Fig. 4).

Notes
1. The given methods describe the use of the NHEJ pathway to knock out a gene by generating random indels. We have adapted these methods to knock in a specific point mutation in IMCD3 cells using the HDR pathway, by integrating methods from the literature [9,10,16]. The basic protocol presented in this chapter can be followed, making the following three changes: (a) In Subheading 3.1, the sgRNA must be designed to target the Cas9 to cut as close to the point at which the single base change is to be introduced, ideally within 10 bp. The necessity of a PAM site being proximal to the desired mutation site is a constraint of the HDR method.
(b) A repair template (single-stranded oligodeoxynucleotide, ssODN) is required. This must contain the desired base change flanked by 60 nt homology arms, as well as a silent PAM mutation. Mutation of the PAM site prevents further Cas9 cleavage after mutation generation. It is important that the PAM mutation is silent to avoid introduction of an unwanted missense mutation. The ssODN should be ordered as an oligonucleotide and resuspended in 2. We have found that HEK293 cells can be transfected with high efficiency using the FuGENE 6 reagent (Promega). In IMCD3 cells, higher transfection efficiency was achieved with the Lipofectamine 2000 transfection reagent (Thermo Fisher) than with FuGENE.
3. If you are not interested in editing a specific locus, the chance of gene knockout will be improved by choosing a sgRNA target site that is in an early exon of the gene. However, it is also worth assessing whether internal translation initiation codons are present within the gene, as these could generate transcripts not affected by a frameshift mutation.
4. Several online sgRNA design tools are now available. We have successfully used CRISPOR (crispor.tefor.net), Chopchop (chopchop.cbu.uib.no), and Benchling (benchling.com) to design sgRNAs. Chopchop does not require the genomic region of interest to be pasted into the search box; instead, the gene of interest can be searched for within the appropriate genome, and all possible sgRNA options spanning the entire gene are presented.

5.
When choosing a sgRNA it is important to consider its score (out of 100). A high score indicates a faithful sgRNA. A low score indicates a sgRNA with predicted off-target activity. For each sgRNA, off-target loci (including those with mismatches) are predicted, which are listed when the sgRNA is selected. The number of off-target sites within genes is also listed. Select a high-scoring sgRNA that does not have any predicted off-target sites in genes.
6. Design primers to generate an amplicon of 200-500 bp. The expected Cas9 cut site (~3 bp upstream of the "NGG" PAM) should not lie centrally within the amplicon, or else two products of roughly equal size will be produced by the T7EI assay and it will not be possible to resolve these by gel electrophoresis. It is recommended that the target genomic region, including the primer-binding sites, is Sanger sequenced within the cell line of choice. Primers should be designed to generate an amplicon that does not contain monoallelic mutations; these will give false positives in the T7EI assay and may bias primer binding (if they fall within primer-binding sites). 7. In a variation of the method presented here, two sgRNAs can be used in combination to excise a genomic region. In this case, two sgRNAs should be designed in Subheading 3.1. Each can be independently cloned into a pX330 vector, to generate two separate pX330-sgRNA plasmids, one for each sgRNA. Transfection conditions can be adapted (in Subheading 3.3), to accommodate transfection of two pX330-sgRNA plasmids with the pAcGFP1-C1 plasmid. Larger deletions can be detected by running PCR products on an agarose gel (Fig. 5).
8. The procedure can be adjusted for other culture vessel sizes; scale reagents as necessary.
9. Confirm GFP expression in CRISPR and GFP-only cells by fluorescence microscopy 18-24 h post-transfection. The GFP-only cells are used purely as a control for transfection, and can be discarded prior to the group sort.  11. It is important to use serum-free medium so as not to clog the FACS apparatus with sera.
12. Cell straining is optional, although it is advised, particularly for IMCD3 cells. This will reduce cell clumps. 14. It is possible to complete the protocol if less than 5 Â 10 5 cells are collected. In this case, prioritize cells for seeding into a culture vessel (Subheading 3.4, step 10); use a vessel of a size appropriate to the number of cells being seeded. If too few cells remain for gDNA extraction (Subheading 3.4, step 11, to Subheading 3.5, step 1), split some of the reseeded cells for gDNA extraction once they have begun to proliferate. 15. We have observed that single-cell sorting cells 2-3 days posttransfection tends to give rise to heterogeneous populations of cells, rather than clonal colonies (they possess more than two alleles when genotyped). We hypothesize that this is due to continued expression of the sgRNA and Cas9 in daughter cells, after division of the single-cell sorted cell. We passage cells for 2 weeks, to await loss of pX330 plasmid from the cells. 16. Optimize PCR conditions according to your needs. Pwo polymerase may be substituted for another polymerase of your choice. Ensure that the polymerase is of high fidelity and/or possesses proofreading ability (3 0 to 5 0 exonuclease activity); errors during PCR amplification will give false positives in the T7EI assay.
17. Use non-transfected cells as a control; these should behave as WT to give a single PCR product which is not cut by the T7EI assay. Multiple PCR products may be resolved from the CRISPR sample if it contains alleles with large indels. If there is a significant size difference between these PCR products, the T7EI assay may produce more than three bands on the gel.
18. sgRNA efficiency can be determined from the intensity of the products resulting from T7EI assay of the CRISPR sample: compare the intensity of the uncut product to that of the cleavage products; the brighter the cleavage products (relative to the uncut product), the more efficient the sgRNA. If T7EI assay of the CRISPR sample does not result in cleavage products, sgRNA activity may be inefficient, and design of a new sgRNA should be considered.
19. Use as many 96-well plates as needed. We typically use 2-3 plates to maximize the number of colonies obtained.
20. It is necessary to passage some cells (of the same line) 24-48 h before preparing transfected cells for FACS, in order to formulate conditioned medium. Conditioned medium is used to improve the likelihood of cell survival after the single-cell sort. It comprises fresh medium (containing necessary nutrients) and medium taken from cells at~70% confluence. The latter contains signaling components secreted by healthy cells in exponential growth phase, to encourage the sorted single cells to grow.
21. Conditioned medium must be sterile-filtered to remove contaminating cells, which may derive from media taken from cells at~70% confluence. Contaminating cells will obstruct the establishment of clonal cell lines.
22. The majority of cells will no longer be GFP positive. Isolate any single cells, regardless of GFP expression.
23. Centrifugation of the 96-well plate improves chances of cell adhesion, by moving cells to the bottom of the plate. It is important to get the single-cell sorted cells into the incubator as quickly as possible after FACS, to maximize viability.
24. Changing medium is not necessary; this risks loss of cells from the wells.
25. It may be necessary to passage the cells before the entire well is confluent, if cells become dense but do not spread across the well, and media become orange/yellow in color.
26. Confirm cell detachment under the microscope prior to transferring cells to the new culture vessel. Transfer the full volume of cell suspension to the 12-well plate without pelleting cells.
Then check the (empty) well of the 96-well plate under the microscope again, to ensure that all cells have been transferred. Loss of cells during passage will result in reduced cell density in the 12-well plate, which may decrease cell viability.
27. Presence of more than two PCR products indicates that the colony is a clone with aneuploidy, or that it is not clonal (it is a heterogeneous population).
28. It is possible to perform the T7EI assay to genotype clones. However, note that T7EI will not cleave PCR products derived from homozygous mutants. T7EI will only give a positive result (cleavage) in PCR products derived from heterozygotes or compound heterozygotes.
29. If Sanger sequencing produces two overlapping traces, the clone is a heterozygote or a compound heterozygote. Cloning of PCR products in pGEM-T Easy vector systems (Promega) can be used to isolate allelic sequences. If more than two overlapping traces are seen, the colony is a clone with aneuploidy, or it is not clonal (it is a heterogeneous population).