Production of isoform-specific knockdown/knockout Madin–Darby canine kidney epithelial cells using CRISPR/Cas9

CRISPR-Cas9 gene editing has made it possible to specifically edit genes in a myriad of target cells. Here, a method for isoform-specific editing and clonal selection in Madin-Darby canine kidney (MDCK) epithelial cells is described in detail. This approach was used to address a long-standing question in virology of how adenovirus enters polarized epithelia from the apical surface. Our method relies on selecting two sgRNA sequences, cloning them into a suitable fluorescently labeled Cas9 vector system, and subsequently transfecting our MDCK epithelium and selecting isoform-specific Coxsackievirus and adenovirus receptor knockout clones. Utilization of this method is readily applicable to many other genetic targets in epithelial cells.• Simultaneous utilization of an sgRNA upstream and an sgRNA downstream of a target sequence allows for deletion of the intervening sequence, including whole exons.• Sorting of cells positive for fluorescent marker gene expression enhances the identification of partial and biallelic gene knockout.• PCR screening allows relatively fast and efficient determination of isoform-specific deletion.


Specifications
(3) Import the validated sequence of your genetic target into the CHOP-CHOP software (chopchop.cbu.uib.no, Harvard University) and select the species from which the target cell line is derived. (4) Use the CHOP-CHOP software to generate a ranked list of potential single guide RNAs (sgRNAs) and select at least one upstream of the target sequence and one downstream of the target sequence. General considerations for sgRNA selection are discussed below.
(1) According to the manufacturer's instruction for Infusion cloning, design primers containing the sgRNA sequence flanked at the 3 end with 23-25 nt of sequence homologous to the plasmid vector. Generate primers using the Infusion Cloning Primer Design Tool ( takarabio.com/ learning-centers/cloning/in-fusion-cloning-tools , Takara Bio USA).  Table 1). (9) Expand two plasmids with confirmed sequences, one which recognizes a sequence upstream of the target and one that recognizes a sequence downstream of the target, by inoculating 250 mL of Luria broth containing 50 mg/ml ampicillin and growing overnight in a 37 °C shaker. (10) Isolate purified plasmid using an EndoFree Plasmid Maxi Kit (Qiagen, Hilden, Germany).
Step 3: Generation and expansion of potential knockout (KO) clones.
(1) Seed MDCK cells at 2.4 × 10 4 cells per well in 24-well plates and grow overnight in MDCK media at 37 °C in a humidified incubator with 5% CO 2 . Seeding at this density allows the cells to be approximately 70% confluent at the time of transfection. Step 4 Screening and validation of potential knockout clones.
(1) Analyze Genomic DNA of potential KO clones by first performing a screening PCR using primer pairs inside the region targeted for deletion with Standard Taq Polymerase (New England Biolabs, Ipswich, MA) using cycling parameters optimized by the manufacturer. The primer sequences that recognized nucleotides within the deleted region used for our JR1-CAR Ex8 -KO cell line were F-GACCCATAAGGGAAGCCTAAC and R-ATGCCTGGTGCCACTTTAT [2] . (2) Subject clones that do not produce an amplicon in the screening PCR to an additional PCR reaction using primers that recognized sequence outside of the expected deletion site. The presence of amplicon size shifts between the parental cell line and potential KO line indicates that DNA deletion events have occurred [2 , 4] . (3) Knockout/knockdown phenotypes can then be validated using a variety of assays. In our case, we measured isoform specific protein expression of CAR and differences in adenoviral transduction [2] . (4) Freeze down aliquots of chosen cells in complete MDCK media with 5% DMSO first at −80 °C prior to being moved to a liquid nitrogen tank for long-term storage. Of note, different cell lines may require more conservative freezing parameters, thus, it is recommended that labs use freezing protocols that have been optimized for their cell line of interest.

Method validation
Step 1: Selection of the single guide RNA sequences (sgRNAs). Simultaneous utilization of two 20 nucleotide sgRNAs, one upstream of the target sequence and one downstream of the target sequence, has previously been described to allow for deletion of the intervening segment [4 , 5] . The rate of diploid deletion events is dependent upon the distance between double stranded breaks with, shorter distances being more efficient [5] . The intronic and exonic sequence for MDCK cells can be obtained from the NCBI canine genome sequence. When designing sgRNA's that recognize intronic sequence, it is important to determine the sequence in the individual cell line that will be used for editing, as discrepancies were noticed between the NCBI sequence and genome of the MDCK cell line used in these experiments. We accomplished this by generating candidate primers, PCR amplifying the regions surrounding the 8th exon of the CXADR gene, and, after PCR product purification (DNA Clean & Concentrator, Zymo Research, Irvine, CA), subjecting these amplicons to commercial Sanger sequencing (Genewiz, South Plainfield, NJ).
Target gene sequences can be entered into the freely available CHOP-CHOP software (Harvard University) to generate candidate sgRNA sequences and rank them based upon predicted cutting efficiency and minimal predicted off target effects [6 , 7] . It is generally considered that sgRNA sequences should be at least 3 nucleotides different than any other sequence present in the genome of the target cell [8] . Furthermore, a guanine nucleotide in the sgRNA immediately adjacent to the PAM sequence has been reported to increase cutting efficiency [9] . We incorporated the sequencing results for the CXADR exon 8 region in our MDCK cell line into the CHOP-CHOP software under the Canis Lupis Familiaris filter and selected a suitable sgRNA that targets intronic sequence upstream of CXADR exon 8 (CGAAGGGCAAAATCTTCTAG) and one that targets intronic sequence downstream CXADR exon 8 (GGTTGCCTTGGGGAAAGTTA).
In order for proper expression of Cas9/sgRNA complexes and selection of transfected cells, a suitable Cas9 expression plasmid must be selected. For sgRNA sequences designed for use in MDCK cells, an Infusion Cloning kit (Takara Bio USA, Mountain View, CA) was used to incorporate sgRNA sequences into the pspCas9(BB) −2A-GFP plasmid (Addgene, Watertown, MA). The main advantage of the pspCas9(BB) −2A-GFP plasmid is that transfected cells turn green, thus making them amenable to FACS to allow enrichment of transfected cells and isolation of potentially edited clonal populations [10] . While this plasmid contains BbSI sites that can allow incorporation of candidate sgRNA sequences using more traditional cloning techniques, performing site directed mutagenesis via inverse PCR with an Infusion cloning kit is another highly efficient viable option for sgRNA integration [10 , 11] .
Step 3: Generation and expansion of potential knockout clones.
After successful production of Cas9 plasmids with properly inserted sgRNA sequences, plasmids must be delivered into the cell line of interest and transfected cells must be isolated for expansion and analysis. We double transfected plasmids containing the sgRNA that recognizes intronic sequences upstream or downstream of CXADR exon 8 into our MDCK cell line using DreamFect Gold transfection reagent (OZ Biosciences, San Diego, CA). Transfection of either plasmid alone resulted in a transfection efficiency of approximately 10-15% ( Fig. 1 ). Two days later, this cell population was subjected to FACS sorting into a 96-well plate, such that a single GFP positive cell was present per well.
Clonal populations were expanded in culture for approximately 2 weeks until wells were near confluence. Despite being careful to ensure adequate nutrition and minimizing the risk of washing single cells off the plate by changing growth media every 3 days, we ultimately expanded 14 clonal populations from a single plate starting with 96 potentially edited cells. Thus, it is important to sort a large number of cells in order to generate a modest number of clonal populations. Furthermore, although not strictly necessary, changing media with a multichannel pipettor speeds up the process and minimizes the time that cells are without media. PCR screening identifies clone 13 as a possible genomic CAR Ex8 knockout clone. A) Schematic representation of the PCR reaction with primers within the region expected to be deleted (PCR 1). If the sequence between the sgRNA that recognizes intronic sequence upstream of CXADR exon 8 (blue arrow) and the sgRNA that recognizes sequence downstream of CXADR exon 8 (orange arrow) is deleted, primers that bind sequence within the deleted segment (red arrows) will be unable to bind and there will be no amplification in the forward (FD) or reverse (RD) direction. B) Deletion PCR reactions for potential CAR Ex8 KO clones (numbers) and MDCK parental cells ( + ) run on an agarose gel. Clone 13 (marked by asterisk) did not amplify, indicating the possibility of a deletion event.
Step 4 Screening and validation of potential knockout clones.
Once DNA is isolated from clonal populations of CRISPR/Cas9 treated cells, screening and validation of the gene-edited phenotype can be performed. It is important to note that cells were selected based upon successful transfection (GFP positivity) and may or may not have undergone gene editing. When Cas9 mediated double stranded breaks occur, there are many potential genetic consequences, such as the incorporation of random insertions and deletions (indels) at the cut site as a result of nonhomolgous DNA repair or reannealing of the broken strand via homologous end joining [12] . Many techniques have been developed to assess the presence of genetic editing [13] . In our method, we wanted to select for cells that had two simultaneous double-stranded breaks on both alleles, one upstream of CXADR exon 8 and one downstream of CXADR exon 8, and deletion of the intervening segment of DNA. In order to screen for such cells, we subjected genomic DNA from each clonal population to two screening PCRs, one with primers that recognize sequence inside of the deleted region (PCR 1) and one that recognize sequence outside of deleted region (PCR 2). This technique allows for the identification and selection of cells that underwent diploid deletion, haploid deletion, or no deletion events [4 , 5] . Of the 14 clones obtained from the first round, one clone (Clone 13) did not amplify a 650 bp band during PCR 1 indicating that it might be a double deletion clone ( Fig. 2 ). This clone and several others were subjected to PCR 2 and a band shift from approximately 1.5 kb in the control cells to 300 bp in clone 13 was seen, suggesting biallelic deletion of the intervening segment containing CXADR exon 8 [2] . This clone was then referred to as JR1-CAR Ex8 -KO. Cell morphology and division rates were then compared between JR1-CAR Ex8 -KO epithelial cells and their parental counterparts. The same number of cells (1.2 × 10 4 cells/well) were seeded in either 24-well plates, for counting, or 4-well chamber slides, for microscopy, and grown for 24, 48, and 72 h in a 37 °C humidified incubator with 5% CO 2 . Representative bright field images were taken at 24, 48, and 72 h post seeding using an Olympus IX83 microscope with an Olympus DP74 microscope digital camera using Olympus cellSens software ( Fig. 3 ). Cells were counted by first lifting with trypsin and then diluting into 1 mL of media. This cell suspension was pipetted vigorously to ensure all cells were lifted and then 10 μL of cell suspension was mixed 1:1 with 0.4% Trypan blue. This mixture was then counted using a Countess II Automated Cell Counter ( Fig. 4 ).
After utilization of screening PCRs to select for double deletion clones, knockout/knockdown of the gene should be confirmed. A variety of techniques can be used for this, including total mRNA isolation and subsequent reverse transcriptase qPCR to measure gene expression and/or methods such as Western blotting or immunohistochemistry to identify expression of the protein of interest. In our case JR1-CAR Ex8 -KO cells were subjected to Western blot with two primary antibodies, one that detects total CAR and one that specifically detects CAR Ex8 , and it was found that cells had drastic reduction in CAR Ex8 expression with only a modest reduction in total CAR expression, consistent with isoform specific knockdown of CAR Ex8 [2] .

Conclusions
The method outlined in this manuscript describes how to use CRISPR/Cas9 technology to delete sections of genomic DNA in MDCK epithelial cells. The deleted segment can be of variable length and if it contains target exons, can result in isoform specific deletion of genes. It is recommended that multiple knockout clones be obtained for any one gene in order to confirm that the resulting phenotype is due to target gene knockout and not potential off-target effects. This method was used to create a virus receptor isoform-specific knockdown cell line by deleting the 8th exon of the CXADR gene. This method is expected to be readily amenable to other genetic targets and to other epithelial cell lines capable of being expanded from single cells.

Declaration of Competing Interest: [MANDATORY -Delete as appropriate]
X The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.