To detect CRISPR-Cas9 nuclease activity in cells, we designed a dual-fluorescence reporter system (pRG2S_Cas9) that facilitates selection of cells expressing the reporter system and cells that have undergone gene editing (Figure 1A). The system constitutively expresses mRFP regardless of any nuclease activity as a marker to enrich cells expressing the reporter construct using flowcytometry. mRFP is followed by a spacer between the mRFP and a stop codon which comprises a Cas9-target sequence where we cloned Cas9:gRNA target sites. The second fluorophore, eGFP, cloned twice out-of-frame (frames +1 and +2) downstream of the stop codon (Figure 1A), allows for the selection of nuclease active cells as previously described (Ramakrishna et al., 2014a). In the presence of functionally active Cas9/sgRNA complexes in the cells delivered as plasmid DNA or RNP, the Cas9 nuclease guided by a sgRNA creates double-stranded breaks in the Cas9-target sequence triggering error-prone Non-homologous End-Joining (NHEJ) DNA repair. This pathway often leads to insertions and deletions (indels), which bypass the stop codon and result in either of the eGFP genes being in-frame and permanently activate eGFP expression in cells (Figure 1B). Thus, the percentage of RFP⁺GFP⁺ cells represents the percentage of cells that have undergone gene editing. This system allows for rapid and effective flow cytometry, microscopy, and microplate reader based detection of nuclease-active cells based on double-positive fluorescence (Figure 1B).

We designed gRNAs towards the inserted Cas9 targeting sequence (Figure 1A) and confirmed their functionality in vitro (Figure 2A). We then compared their efficiency in HEK293 cells co-transfected with the reporter plasmid, a CRISPR-Cas9 expression plasmid, and either gRNA expression fragments (Figure 2B). The highest nuclease activity was attributed to target reporter-gRNA2 based on the resultant percentage of GFP expressing cells (Figure 2B).

Having determined that the dual reporter construct can be used to detect Cas9 nuclease activity, it was stably incorporated into HEK293 cells that were sorted into low and high expressing populations (Figure 2C and Supplementary Figure S1A). As expected, stable reporter HEK293 cells showed a high expression of mRFP, and only showed expression of eGFP after transfection with plasmids expressing Cas9 and targeting gRNAs (Figures 2C,D and Supplementary Figure S1C). Cas9 target reporter-gRNA2 was again found to be the most efficient in stable reporter cells co-transfected with a CRISPR-Cas9 expression plasmid, and either of the four gRNA expression fragments (Supplementary Figure S1B). The gRNA with the highest efficiency, Cas9 target reporter-gRNA2, was thus selected for further experiments based on the highest GFP expression in cells as measured by flow cytometry (Supplementary Figure S1B). gRNA2 was cloned into the CRISPR-Cas9 expression plasmid, then used as the single plasmid delivery cargo for subsequent experiments. For ribonucleoprotein (RNP) based assays, a Cas9 nuclease complexed with a synthetic gRNA harboring the sequence of Cas9 target reporter-gRNA2 was used.

To enably high throughput evaluation of CRISPR delivery efficiency and nuclease activity, we developed a microplate reader based protocol to directly and rapidly quantify CRISPR nuclease activity in cells using the dual-fluorescent stable reporter system (Figure 3, Supplementary Figure S2). We utilized the highly sensitive FLUOstar^® Omega multichromatic microplate reader that enables low signal detection levels. We seeded RFP⁺GFP⁺ cells at defined numbers and initial optimization results indicated that the FLUOstar^® Omega microplate reader was able to detect a distinct GFP signal at all cell numbers (Supplementary Figure S2A). Results showed a linear relationship between GFP/RFP fluorescence intensity ratio and seeded RFP⁺GFP⁺ cell numbers (Supplementary Figure S2B). This relationship was confirmed by counting RFP⁺GFP⁺ cells using flowcytometry (Supplementary Figure S2C. We were thus able to generate a formula from which the percentage of nuclease active RFP⁺GFP⁺ cells can be determined from the microplate reader’s GFP/RFP fluorescence intensity output (Supplementary Figure S2D).

To demonstrate the applicability of this approach, we transfected stable reporter HEK293 cells with different amounts of Cas9/gRNA2 RNP using Lipofectamine CRISPRMAX and measured nuclease activity at 48 and 72 h using the microplate reader (Figures 3B,C). As expected, different levels of GFP/RFP signals were detected at each RNP amount and increased over time (Figure 3B). Interestingly, the percentages of RFP⁺GFP⁺ cells obtained using flowcytometry were highly similar to the values we calculated using the formula generated above (Figure 3D), further verifying the microplate reader approach as an efficient method for measuring the levels of CRISPR nuclease activity using the dual-fluorescent stable reporter system.

In order to further establish our fluorescent stable reporter system as a means for the enrichment of CRISPR-induced mutations in endogenous genes, we transfected stable fluorescent reporter HEK293 cells with CRISPR plasmid or RNP carrying gRNAs targeting the reporter Cas9-target sequence (Cas9-reporter gRNA2) and endogenous genes simultaneously. As a proof-of-concept, we chose to target SYNE4, a gene that encodes nuclear membrane protein nesprin-4 and is associated with hearing loss (Horn et al., 2013), EMX1 associated with neurodevelopmental and infertility disorders (Zhang et al., 2019), (Kim et al., 2010), and SNCA associated with neurodegenerative Parkinson’s Disease (Spillantini et al., 1997), (Bonini and BI, 2005). The percentage of RFP⁺GFP⁺ cells obtained in RNP (Cas9/gSNCA) targeted cells (∼16%) (Supplementary Figure S3A) was similar to the percentage we obtained above (Figure 3D). After gene targeting, we sorted pureRFP⁺GFP⁺ cells using fluorescence-assisted cell sorting (Figures 4Ai,Bi). Analysis of indels in RFP⁺GFP⁺ using T7 Endonuclease I (T7EI) assay (Figures 4Aii,Bii revealed significant disruption of endogenous genes in GFP expressing cells using plasmid (targeting SYNE4 gene: 60%; targeting EMX1 gene: 56%; targeting SYNE4 and EMX1 genes simultaneously: 33 and 36%, respectively) (Figure 4Aii) or RNP (SNCA gene: 42%) (Figure 4Bii). The formation of indels in RFP⁺GFP⁺ cells was further confirmed by Sanger’s sequencing of PCR clones (Figures 4Aiii,Biii) and the frequency of indels was analysed by Tracking of Indels by Decomposition (TIDE) assay (Supplementary Figure S4).

To demonstrate the applicability of our fluorescent reporter system to other applications and cell types, we employed the same strategy to knockout a protein-coding gene in neuronal SH-SY5Y cells (Figure 5). We chose to target SNCA, a gene abundantly expressed in the brain and associated with Parkinson’s disease (Spillantini et al., 1997), (Bonini and BI, 2005). After evaluating gRNAs targeting exon 4 of SNCA, we identified SNCA-gRNA3 as the most efficient gRNA for reducing the expression of endogenous SNCA protein in SH-SY5Y cells (Supplementary Figure S5). We then transfected stable reporter SH-SY5Y cells with CRISPR targeting the reporter sequence (Cas9-reporter gRNA2) and SNCA gene using SNCA-gRNA3, simultaneously, and sorted the treated cell pool into RFP⁺GFP⁻ and RFP⁺GFP⁺ cells (Figure 5A and Supplementary Figure S3B) and analyzed sorted cells using T7EI assay and western blot (Figures 5C,D and Supplementary Figure S3C). We found significant levels of gene disruption in RFP⁺GFP⁺ cells (56%) compared with RFP⁺GFP⁻ cells (8%) as shown by T7EI assay (Figure 5C). Similarly, RFP⁺GFP⁺ stable reporter SH-SY5Y cells exhibited a significant diminishment of SNCA protein expression (18% SNCA expression) compared with RFP⁺GFP⁻ (84% SNCA expression) relative to untransfected group as shown by western blot (Figure 5D). This diminished expression is in agreement with Sanger’s sequencing data of PCR clones revealing the formation of premature stop codons in 67% of sequenced clones, possibly knocking out protein expression via the process of non-sense mediated decay (Popp and Maquat, 2016) (Figure 5B). Furthermore, sequencing analysis of RNP treated cells revealed the formation of deletions at the target site in 73% of PCR clones (Supplementary Figure S3D).

It is well reported that MPP⁺, the toxic metabolite of the neurotoxin MPTP, induces cell death in neuronal cells expressing SNCA (Dauer et al., 2002; Fountaine et al., 2011; Javed et al., 2016). To evaluate a functional role for SNCA, we tested response of RFP⁺GFP⁺ SNCA knock-out cells to MPP⁺ and detected a significant reduction in MPP⁺-mediated cytotoxicity compared with the wildtype reporter SH-SY5Y cells (Figure 5E). Thus, SNCA is important for mediating the response to the neurotoxin MPTP and knocking out SNCA expression protects neuronal cells from MPP⁺-induced cytotoxicity. These results validate that our fluorescent stable reporter system can be used to enrich cells carrying mutations in multiple endogenous genes and suggest it may facilitate the generation of knockout cell lines useful for downstream applications such as studying protein function.

gRNAs were designed using Benchling’s online gRNA design tool (see Supplementary Data for sequences) and were synthesized as tracrRNA and crRNA and purchased from Integrated DNA Technologies (IDT Coralville, IA, United States) or as sgRNA (Synthego, CA, United States) or embedded in gRNA expression fragments under a U6 promoter (IDT Inc.). gRNAs were designed using benchling’s online gRNA design tool and purchased from Synthego, and gSNCA4 was purchased from IDT Inc.

To construct the pCas9/gRNA2 plasmid, gRNA2 from U6-pRG2S-Cas9-gRNA2 fragment was cloned into the gRNA scaffold of a Cas9 expression plasmid, pX330-U6-Chimeric_BB-CBh-hSpCas9, a gift from Feng Zhang (Addgene plasmid #42230) by double digestion using BbsI purchased from New England Biolabs (NEB, MA 01938, United States). One Shot DH5-α competent E. coli cells (Invitrogen, Thermo Fisher Scientific, MA 02451, United States) were transformed with the ligation reaction and plated onto LB agar plates containing 100 μg/ml of Ampicillin (Thermo Fisher Scientific, MA, United States).

pCas9/gSYNE4, pCas9/gEMX1 and pCas9/gSNCA3 plasmids were constructed by adding the gRNA sequence into the Cas9 expression vector using overlapping oligos. Once annealed, the oligos produce a fragment containing the gRNA sequence flanked by BbsI overhangs and are cloned by ligation into a Cas9 expression vector linearized by BbsI. One Shot DH5-α competent E. coli cells were transformed with the ligation reaction and plated onto LB agar plates containing 100 μg/ml of Ampicillin (Thermo Fisher Scientific).

To construct the reporter plasmids (pMRS_Cas9: Catalog #TGEN_dRRR1; pRG2S_Cas9: Catalog#TGEN_dRRM1, ToolGen, Seoul, South Korea), we designed a DNA sequence containing multiple CRISPR-Cas9 gRNA binding sites flanked by BstXI and BamHI cloning sites. The sequence harboring the Cas9 gRNA bindings sites was synthesized by IDT and purchased as a pIDT-plasmid. The targeting sequence was sub-cloned from pIDT into a the reporter plasmids between the mRFP and EGFP fluorescent reporter genes by double digestion using BstXI and BamHI (NEB). One Shot DH5-α competent E. coli cells were transformed with the ligation reaction and plated onto LB agar plates containing 50 μg/ml of Kanamycin (Thermo Fisher Scientific). Positive clones were verified by Sanger’s sequencing using pRG2S forward and reverse primers, synthesized by IDT Inc. (see Supplementary Data for sequences).

The Ribonucleoprotein (RNP) complex used in the vitro study was prepared in similar manner described by Saifaldeen et al. (Saifaldeen et al., 2021). Briefly, a final concentration of 50 nM Cas9 nuclease was added to a NEB Buffered solution containing 300 ng pRG2S_Cas9 plasmid and 50 ng gRNA, topped up to 20 μl filter sterilized distilled deionized nuclease-free water. The reaction was allowed to take place at 37°C for 1 h. Results were analyzed on an Ethidium Bromide-stained 0.8% agarose gel using a UV transilluminator (ChemiDoc™ Imaging System, Biorad, CA, United States).

Human Embryonic Kidney cells (HEK293) were maintained in high glucose Dulbecco’s modified Eagle’s medium (DMEM, high glucose, Catalog# Gibco™11965092, Thermo Fisher Scientific) supplemented with 10% foetal bovine serum (Catalog# Gibco™10082147, Thermo Fisher Scientific) 1% penicillin and streptomycin (Catalog# Gibco™15070063, Thermo Fisher Scientific). Neuronal SH-SY5Y cells were maintained in Dulbecco’s Modified Eagle Medium/Nutrient Mixture F-​12 (DMEM/F12 1:1, Catalog# Cytiva HyClone™SH30272.02, Thermo Fisher Scientific) supplemented with 20 mM HEPES buffer (Catalog# Gibco™ 15630080, Thermo Fisher Scientific)), 10% FBS, and 1% penicillin and streptomycin. Cells were grown in an incubator at 37°C in 5% CO₂ and 95% humidity.

To generate a stable reporter cell line, cells were transfected with 10 μg BsaI-linearized pRG2S_Cas9 plasmid (Figure 1A) using FuGENE (Catalog# E2311, Promega, WI, United States) (3 μl/μg). After 3 days, Geniticin G418 Sulphate (Catalog# Gibco™ 0131035, Thermo Fisher Scientific) was added (500 μg/ml for HEK293, 300 μg/ml for SH-SY5Y cells) to select stable cells over a period of 2 weeks. A high (∼98% RFP) and low (∼72% RFP) expressing stable poopulations was sorted using a FACS sorter (BD FacsAria™ II Cell Sorter, BD Biosciences, CA, United States). Fluorescence microscopy was used to confirm RFP expression over several passages.

Choosing the best performing gRNA: HEK293 cells were seeded at 8 × 10⁴ cells per well in 500 ul complete DMEM on a 24-well plate (Catalog# Nunc™ 142475, Thermo Fisher Scientific). After 24 h, cells were co-transfected with 100 ng pMRS_Cas9 reporter plasmid, 150 ng U6-gRNA fragment of different gRNAs, and 500 ng pX330-U6-Chimeric_BB-CBh-hSpCas9 Cas9 expression plasmid using FuGENE (3 μl/μg following manufacturer’s instructions) and incubated for an additional 72 h. At the day of analysis, cells were detached using TrypLE (Catalog# Gibco™12605010, Thermo Fisher Scientific) and collected in DMEM; GFP expression and fluorescence intensity was evaluated using a flow cytometer (LSRFortessa, BD Biosciences).

Comparing gRNA activity in stable cells: HEK293 cells were seeded at 15,000 cells per well in 100 μl complete DMEM on a 96-well plate (Catalog# Nunc™167008, Thermo Fisher Scientific). The next day, cells were co-transfected with 30 ng U6-gRNA fragment of different gRNAs, and 100 ng pX330-U6-Chimeric_BB-CBh-hSpCas9 Cas9 expression plasmid using FuGENE (3 μl/μg). Cells were analyzed after 72 h using flowcytometry.

Microplate reader assay: stable reporter HEK293 cells were seeded at 10,000 per well in 100 µl phenol-free Opti-MEM the day before transfection. Wildtype, non-fluorescent HEK293 cells were seeded at the same density for correction. Cells were then transfected with different amounts of RNP as indicated using Lipofectamine CRISPRMAX (Catalog# CMAX00001, Thermo Fisher Scientific) following the manufacturer’s protocol. Cells were topped-up with 100 μl phenol-free Opti-MEM after 6 h.

Endogenous gene targeting (HEK293): stable reporter cells were seeded at 25,000 cells per well in 500 μl Opti-MEM on a 24-well plate. The next day, cells were transfected with a total of 1 μg pCas9/SYNE4 and/or pCas9/EMX1 plasmids using FuGENE (3 μl/μg). For RNP transfection, cells were transfected with RNP containing gRNA #3 targeting SNCA gene (Synthego) using Lipofectamine CRISPRMAX following manufacturer’s protocol.

Endogenous gene targeting (SH-SY5Y): stable repoter cells were seeded at 37,500 cells per well in 500 μl OptiMEM on a 24-well plate. The next day, cells were transfected with a total of 1 μg pCas9/gSNCA3 plasmid using Lipofectamine 2000 as per manufacturer’s protocol. For RNP transfection, cells were transfected with RNP containing gRNA #3 targeting SNCA gene using Lipofectamine RNAiMAX following manufacturer’s protocol.

Cells were topped up with 20% FBS containing Opti-MEM after 6 h. Cells were grown until they reached confluency and then transferred to a 6-well plate (Catalog# Nunc™145380, Thermo Fisher Scientific) and grown in complete DMEM until they reached confluency. Cells were further transferred to a T75 flask (Catalog # Nunc™156499, Thermo Fisher Scientific) before sorting.

To determine whether the microplate reader is sensitive enough to detect differences in cellular GFP expression, we seeded RFP⁺GFP⁺ expressing cells at defined cell numbers (1,000, 5,000, 10,000, 15,000, 20,000, and 30,000) in a 96-well plate (Catalog# 164588, Thermo Scientific Nunc Optical CVG, black-well, clear bottom) in a phenol-free medium to minimize fluorescence background (Supplementary Figure S2C). Cells were diluted with RFP⁺GFP⁻ cells to maintain a constant cell number of 40,000 per well and the plate was placed in the incubator overnight. 40,000 RFP⁻GFP^- were seeded and used as blank controls.

Fluorescence was measured using the filter-based multi-mode FLUOstar^® Omega microplate reader. Bottom optical scan was performed using Spiral Averaging or 10 × 10 Matrix Scan (6 mm diameter wells) to eliminate uneven cell distribution. Gain was determined using the microplate reader’s built-in gain adjustment option (RFP; 3,468, GFP; 2,259). Multiple scans were run to determine the optimal excitation and emission wavelengths (RFP; λ_EX—584 nm, λ_EX—620-10 nm, and GFP; λ_EX—485-12 nm, λ_EX—520 nm) based on signal to background ratio. Fluorescence intensities were corrected using signals obtained from RFP⁻GFP^- cells, and GFP/RFP ratios were calculated from normalized values.

By plotting normalized GFP/RFP ratios against % RFP⁺GFP⁺ cells counted using flowcytometry, the following formula was generated from the linear trendline: % R F P + G F P + = 3 ( G F P / R F P ) 5 × 100

To test and compare the assay with conventional flowcytometry, cells were transfected as described in Section 4.6. At the timepoints indicated in results Section 2.3, the plate was sealed using a transparent microplate sealer and read as described above. At 72 h, cells were additionally collected and analyzed using flowcytometry as described below.

HEK293 cells were dissociated by trypsinization for 5 min at 37°C and collected. Cells are filtered using a 50 μm cut-off cell strainer (BD Biosciences) to remove aggregates and collected in 5 ml round-bottom collection tubes. Flow cytometry was carried out in a LSRFortessa (BD Biosciences) machine, analyzing 10,000 events per sample for PE-CF594 (RFP) and Alexa-Fluor 488 (GFP) expression (Supplementary Figure S6A). The percentage of editing was then deduced by calculating the percentage of RFP⁺/GFP⁺ from all RFP⁺ cells.

Cell Sorting: After treatment with gRNA’s targeting the reporter gene and endogenous genes simultaneously, RFP⁺GFP⁺ and RFP⁺GFP⁻ cells were sorted using BD FACSAria II Cell Sorter (Supplementary Figure S6B). Cells were grown separately in 96-well plate and transferred to a 24-well plate and then to a 6-well plate. At confluency, cells were detached for DNA and protein analyses.

Genomic DNA was isolated from treated and untreated cells as control using QIAamp DNA Mini Kit (Catalog# 51306, Qiagen, Hilden, Germany) following the manufacturer’s protocol and protocol used by Aouida et al., 2015 (Aouida et al., 2015). PCR was carried out to amplify SYNE4 target using AccuPrime High Fidelity Taq Polymerase (Catalog# 12346-086, Invitrogen, CA, United States), according to the manufacturer’s instructions for 40 cycles (94°C, 30 s; 65°C, 30 s; 68°C, 60 s). EMX1 target was amplified using AccuPrime High Fidelity Taq Polymerase (annealing at 57°C). To amplify SNCA target, Phusion Hot Start Flex 2X Master Mix (NEB) was used, according to manufacturer’s instructions for 30 cycles (98°C, 30 s; 65°C, 35 s; 72°C, 30 s). Each mix was vortexed and centrifuged before placing into a thermal cycler. To confirm PCR, 2 μl of reaction mixed with 1ul loading dye (6X) and 3 μl H₂O were loaded and run on a 2% agarose (Catalog# 16500100, Thermo Fisher Scientific) gel. Only one band of the expected size should show per sample. PCR products are then purified and eluted in water using Qiagen’s PCR purification kit manufacturer’s protocol.

For measuring SNCA protein expression after enrichment of cells containing SNCA mutations, cells were suspended in lysis buffer (MT Cell Lysis Reagent, Catalog# CelLytic C3228-50ML, Sigma) containing 10 μl/ml Protease and Phosphatase Inhibitor Cocktail (Catalog# Halt™ 78440, Thermo Fisher Scientific) and lysed using manufacturer’s protocol. 30 μg cell lysates were loaded on 15% SDS polyacrylamide1 mm gel and separate by electrophoresis at 100 V for 90 min. Proteins were subsequently transferred onto a nitrocellulose membrane performed at 100 V for 60 min. The membrane was immediately placed in boiling PBS for 5 min, blocked with 5% blocking buffer (Blotting-grade blocker, Catalog#1706404, Biorad) in PBS-T for 1 h then incubated with 1:2,500 anti-α-synuclein antibody (Syn-1, Catalog# 610787, BD BioSciences) and β-actin antibody (ab8227, abcam) overnight at 4°C. The protein bands were visualized using Super Signal West Femto Chemiluminescent substrate (Thermo Scientific) and the band intensities determined using Quantity One-4.1.1 software (Bio-Rad) and ImageJ software (NIH, Bethesda, MD).

To assess MPP + toxicity, we measured the cellular redox activity with MTT as previously described (Javed et al., 2016). Briefly, reporter SH-SY5Y cells were plated at a density of 5,000 cells per well on a 96-well plate in 100 µl complete media. The next day, cells were exposed to 1-methyl-4-phenyl pyridinium (MPP + iodide, Catalog# D048-100 MG, Sigma) for 24 h. MTT reagent (Catalog# M2128-1G, Sigma) were then added to a final concentration of 0.6 mg/ml and incubated for 4 h. Media was removed and 100 µl of cell lysis buffer, composed of15% SDS (Catalog# 11667289001, Sigma), 50% N,N-dimethylformamide (Catalog# 227056-100ML, Sigma), pH 4.7, was then added to the cells and incubated in a humidified incubator overnight at 37°C. Absorbance was then measured at 590 nm using a plate reader (TECAN).