CIRCLE-Seq for Interrogation of Off-Target Gene Editing

Résumé

A significant barrier to technologies like CRISPR is the off-target events that can disrupt vital genes. 'Circularization for In Vitro Reporting of Cleavage Effects by Sequencing' (CIRCLE-seq) is a technique designed to identify unintended cleavage sites. This method maps the genome-wide activity of CRISPR-Cas9 with high sensitivity and without bias.

Résumé

Circularization for In Vitro Reporting of Cleavage Effects by Sequencing (CIRCLE-seq) is a novel technique developed for the impartial identification of unintended cleavage sites of CRISPR-Cas9 through targeted sequencing of CRISPR-Cas9 cleaved DNA. The protocol involves circularizing genomic DNA (gDNA), which is subsequently treated with the Cas9 protein and a guide RNA (gRNA) of interest. Following treatment, the cleaved DNA is purified and prepared as a library for Illumina sequencing. The sequencing process generates paired-end reads, offering comprehensive data on each cleavage site. CIRCLE-seq provides several advantages over other in vitro methods, including minimal sequencing depth requirements, low background, and high enrichment for Cas9-cleaved gDNA. These advantages enhance sensitivity in identifying both intended and unintended cleavage events. This study provides a comprehensive, step-by-step procedure for examining the off-target activity of CRISPR-Cas9 using CIRCLE-seq. As an example, this protocol is validated by mapping genome-wide unintended cleavage sites of CRISPR-Cas9 during the modification of the AAVS1 locus. The entire CIRCLE-seq process can be completed in two weeks, allowing sufficient time for cell growth, DNA purification, library preparation, and Illumina sequencing. The input of sequencing data into the CIRCLE-seq pipeline facilitates streamlined interpretation and analysis of cleavage sites.

Introduction

Genome engineering has seen significant advancements over the past twenty years, with a major milestone being the discovery of clustered regularly interspaced short palindromic repeats (CRISPR)-Cas9 in 2012¹. Leveraging the programmable nature of bacterial DNA endonucleases, CRISPR-Cas9 technology enables precise targeting and modification of almost any DNA sequence. Since its inception, the system has been optimized to rely only on the Cas9 endonuclease and a guide RNA (gRNA) to edit specific genomic regions. CRISPR-Cas9's potential as a curative therapy has been demonstrated in clinical trials for various conditions such as Leber's congenital amaurosis, transthyretin amyloidosis, and sickle cell anemia, among others^2,3,4.

CRISPR-Cas9 induces double-stranded breaks (DSBs), which are typically resolved by one of two mechanisms: the error-prone non-homologous end joining (NHEJ) or the more precise homology-directed repair (HDR), provided a template DNA is available. The tendency of CRISPR-Cas9 to cause NHEJ-associated insertions and deletions (indels), along with cleavage at unintended genomic sites, limits its application in clinical settings^5,6,7,8,9,10. Additionally, unintended genomic modifications can create cryptic splice sites, nonsense or missense mutations, induce chromothripsis, or confer oncogenic potential to cells-outcomes that have been observed in several genome editing trials^{11,12,13,14,15}. In conclusion, accurately identifying the off-target activity of CRISPR-Cas9 is crucial for its clinical applications, particularly in systemic gene therapies that may alter billions of cells.

Various methods can be employed to identify CRISPR-Cas9 off-target cleavage sites, including Genome-wide Unbiased Identification of Double-stranded breaks (GUIDE)-seq¹⁶, which uses double-stranded oligodeoxynucleotides to tag DSBs in living cells. Nevertheless, a criticism of this method is that false positives can arise from random DSBs or from PCR artifacts, which must be discarded by excluding captured sites that show poor similarity to the on-target sites. The method based on the use of Integrase-Defective Lentiviral Vector (IDLV) is less sensitive and likely to miss many off-target sites¹⁷. Other in situ methods like DSBCapture, BLESS, and BLISS^18,19,20 involve fixed cells and label DSBs directly, however they are constrained by their dependency on immediate DSB capture and the absence of exogenous DNA. Digenome-seq²¹, an in vitro method, and Selective enrichment and Identification of Tagged genomic DNA Ends by sequencing (SITE-seq)²² both provide sequencing solutions but have their limitations in background noise and single-end analysis, respectively. Discovery of in situ Cas off-targets and verification by sequencing (Discover-Seq)²³ offers in vivo and in situ identification of Cas9 activity via MRE11 binding, but only detects DSBs that exist at the time of sample preparation²⁴. Lastly, Inference of CRISPR Edits (ICE) uses a bioinformatics approach to robustly analyze CRISPR edits using Sanger data²⁵.

This article describes a detailed procedure for Circularization for In Vitro Reporting of Cleavage Effects by Sequencing (CIRCLE-seq): an in vitro technique that sensitively and impartially maps the genome-wide off-target activity of Cas9 nuclease in a complex with the gRNA of interest²⁶. This approach begins with culturing the cells of interest and isolating DNA, followed by random shearing through focused ultrasonication, then exonuclease and ligase treatment. This process ultimately produces circular double-stranded DNA molecules, which are then purified through plasmid-safe DNase treatment. This circular DNA is then exposed to the Cas9-gRNA complex, which cleaves at both intended and unintended cleavage sites, leaving behind exposed DNA ends that act as substrates for Illumina adapter ligation. This process produces a diverse library of genomic DNA (gDNA) containing both ends of each nuclease-induced DSB, ensuring that each read has all the information necessary for each cleavage site. This allows for the use of Illumina sequencing with lower sequencing coverage requirements, setting CIRCLE-seq apart from other similar methods mentioned above. It is important to note that while CIRCLE-seq does have higher off-target sensitivity than other protocols as an in vitro method, this comes at the cost of higher false-positives due to the absence of the epigenetic landscape that is present in other methods such as GUIDE-seq¹⁶. Additionally, DSB DNA repair and its associated machinery are not present in CIRCLE-seq, abrogating indels or proper repair that would otherwise be observed.

In addition to describing the step-by-step protocol to perform CIRCLE-seq, the protocol is validated by identifying genome-wide unintended cleavage sites of CRISPR-Cas9 that occur during the modification of the AAVS1 locus, as an example. This easy-to-follow protocol provides detailed instructions, from the culture of induced pluripotent stem cells (iPSCs) and gDNA isolation to gDNA circularization, Cas9-gRNA cleavage, library preparation, sequencing, and pipeline analysis. Given the low sequencing coverage requirements, CIRCLE-seq is available to any lab with access to next-generation sequencing.

Protocole

The details of the reagents, consumables, and equipment used for this study are listed in the Table of Materials.

1. Cell culture (5 days)

1. Include a negative control throughout this protocol. Prepare enough cells for an additional 25 μg of gDNA (~2.0e⁷ cells per sample).
2. Culture iPSCs according to the established protocol²⁷. Collect the cells and resuspend them in 10 mL of PBS. Pipette a 6 μL sample of cells and resuspend in a 1:1 ratio with trypan blue. Use automated cell counter to count the sample.
3. Aliquot 2 x 10⁷ cells per tube, then spin down at 300 x g for 3 min at 25 °C (Room temperature (RT)). This is sufficient for multiple replicates. Pipette off and discard the supernatant.

2. Genomic DNA isolation (1 day)

1. Use the commercially available DNA purification kit to isolate the gDNA, following the manufacturer's instructions:
   1. Add 200 µL of PBS to a 15 mL conical tube containing the cell pellet and resuspend. Then, pipette 3 mL of cell lysis buffer and 15 µL of proteinase K into the tube. Invert the tube 25 times to mix thoroughly. Place the tube in a water bath shaker set to 55 °C and 150 rpm for 3 h, or overnight for optimal DNA yield.
   2. Pipette in 15 μL of RNase A. Invert 25 times. Place in a water bath at 37 °C for 1 h.
   3. Chill the sample on ice for 5 min. Then, add 1 mL of protein precipitation solution, vortex at high speed for 20 s, and centrifuge at 2000 x g for 10 min at RT. Proteins should form a visible, compact pellet at the bottom of the tube. If the pellet is not visible, incubate the sample on ice for an additional 5 min and centrifuge again.
   4. Add 3 mL of 100% isopropanol to a new 15 mL conical tube. Carefully pipette the supernatant from step 2.1.3 into the tube. Invert the tube 50 times to mix, then centrifuge at 2000 x g and room temperature for 3 min. Without disturbing the DNA pellet, carefully aspirate the supernatant using a Pasteur pipette connected to a vacuum trap and invert the tube on a clean, lint-free wipe.
   5. Pipette 3 mL of 70% (v/v) ethanol onto the DNA pellet and invert 10 times to wash. Centrifuge at 2,000 x g for 3 min at RT. Then, carefully pour off the supernatant.
   6. Keep the tube open and allow the resulting DNA pellet to dry for 30 min, ensuring that all ethanol has fully evaporated. Pipette in 50 μL of DNA hydration solution and mix thoroughly by gentle pipetting.
   7. Dissolve the DNA by placing the sample in a water bath shaker at 65 °C for 1 h, then leave the sample out at RT overnight. Centrifuge the sample at RT and 2,000 x g for 1 min and use the dsDNA BR Assay Kit with associated tubes to quantify the isolated DNA with the fluorometer.

3. Preparation of gRNA (7 days)

1. Order the synthetic gRNA of interest from a commercial source (see Table of Materials). This protocol is also compatible with crRNA/tracrRNA.

4. gRNA in vitro cleavage test

NOTE: Here, a target in the AAVS1 gene is used. To target other genes of interest, design primers (Table 1) to amplify the target region and replace the primers in the following steps with custom primers.

1. Prepare the PCR reaction: Mix 25 μL of Phusion Hot Start Flex 2x Master Mix (Final Concentration 1x), 0.5 μL of AAVS1 F Primer (Final concentration 0.1 μM), 0.5 μL of AAVS1 R Primer (Final concentration 0.1 μM), 5 μL of gDNA (100 ng, 20 ng/μL, from step 2.1.6), and 19 μL of Nuclease-free H₂O (Total volume: 50 μL).
2. Use the following thermocycler parameters: Denaturation: 98 °C for 2 min (1 cycle), Denaturation: 98 °C for 10 s (10 cycles), Annealing: 72-62 °C (-1 °C/cycle) for 15 s (10 cycles), Extension: 72 °C for 30 s (10 cycles), Denaturation, 98 °C for 10 s (30 cycles). Annealing: 65 °C for 15 s (30 cycles), Extension: 72 °C for 30 s (30 cycles), Final Extension: 72 °C for 5 min (1 cycle), Hold: 4 °C indefinitely.
3. Use AMPure XP beads to purify the product of the PCR reaction. First, pipette in 1.8x volumes, or 90 μL, of XP beads to the PCR product. Pipette ten times to thoroughly mix. Leave the mixture at RT for 5 min to incubate.
4. Using a magnetic rack, separate the beads from the solution by placing the PCR reaction plate on the magnet for 3 min. Pipette off the cleared solution and discard. Add 200 µL of 80% ethanol (v/v) to the beads, incubate for 30 s, and then remove the ethanol. Repeat this washing step twice to ensure complete removal of the ethanol.
5. Let the samples dry naturally for 3 min by placing the plate on the magnet. Remove the plate from the magnet and add 40 µL of TE buffer, pH 8.0. Mix by pipetting up and down ten times. Allow the sample to sit at RT for 2 min.
6. Place the PCR reaction plate on the magnet for an additional minute. After 1 min, transfer the supernatant to a new plate. Measure the purified PCR yield using a spectrophotometer, and analyze it on a TapeStation using an optical tube strip with an optical tube strip cap, along with High Sensitivity D1000 ScreenTape and High Sensitivity D1000 reagents (ladder and buffer), following the manufacturer's instructions. Store the prepared sample at -20 °C for up to several months.
7. Dilute Cas9 nuclease protein to 1 μM as follows: Mix 2 μL of 10x Cas9 buffer (Final concentration 1x), 1 μL of Cas9 Nuclease, S. pyogenes (Final Concentration 1 μM), and 17 μL of Nuclease-free H₂O in a total volume of 20 μL.
8. Perform an RNase-free procedure to prevent gRNA degradation. Dilute gRNA (from step 3.1) to 3 μM in H₂O to a total volume of 10 μL.
   NOTE: Use the following formula to estimate the molecular weight of gRNA: Molecular weight of ssRNA (g/mol) = (length of ssRNA (nt) x 321.47 g/mol) + 18.02 g/mol. For reference, a 104 nt long gRNA at 3 µM is approximately 100 ng/µL.

5. DNA shearing (3 h)

1. Prepare the ME220 by first homing the control arm. Then, fill the reservoir with purified deionized H₂O. On the control station laptop, access Water Works and click on Fill. Adjust the temperature to 4.5 °C.
2. Transfer 25 μg of gDNA into a microtube (microtube-130 AFA Fiber Screw-Cap). Then, fill the tube to a total volume of 130 μL with 1x TE. Use the following conditions to shear the DNA to an average length of approx. 300 bp: set Duration to 10 s; Peak Power to 70; Duty Factor % to 20; Cycles/Burst to 50; all of which will automatically set the Avg Power to 14.0.

6. Purifying sheared genomic DNA (1 h)

1. Divide the sheared genomic DNA into two portions of 65 µL each. Purify using 1.8 times the volume of XP beads (117 µL), following the procedure outlined in steps 4.3-4.6. Transfer the supernatant to a new PCR plate and measure the quantity using a spectrophotometer.
2. Run 1 μL of the eluted sheared gDNA on a TapeStation, according to the manufacturer's instructions, to ensure that the gDNA is sheared to a broad distribution of approx. 300 bp. If needed, store the sheared gDNA at -20 °C for up to several months.

7. Preparation of CIRCLE-seq library (3 days)

1. Hairpin adapter annealing
   1. Resuspend oSQT1288 (Table 1), the hairpin adapter, to a final concentration of 100 μM in 1x TE.
   2. Perform the adapter annealing as follows: mix 40 μL of oSQT1288 (Final Concentration 40 μM), 10 μL of 10x STE (Final Concentration 1x), and 50 μL of Nuclease-Free H₂O for a total volume of 100 μL.
   3. Use the following the annealing parameters: 95 °C for 5 min, -1 °C per min for 70 cycles, hold at 4 °C indefinitely.
2. Perform End repair . Employ the PCR-Free HTP Library Preparation Kit and prepare the End-Repair Master Mix.
   1. Mix 8 μL of Nuclease-Free H₂O, 7 μL of 10x end-repair buffer (Final Concentration 1x), and 5 μL of end-repair enzyme mix (final volume 20 μL of Total End-Repair Master Mix).
   2. Pipette 20 μL of the end-repair master mix into the sheared gDNA sample from steps 4.3-4.6. Mix 20 μL of End-Repair Master Mix with 50 μL of sheared gDNA for a final volume of 70 μL.
   3. Place the mix in a thermocycler at 20 °C for 30 min, then hold at 4 °C indefinitely.
   4. Add 1.7x volumes, or 120 μL, of XP beads and follow the purification steps from steps 4.3-4.6. Elute with 42 μL of TE, pH 8.0. Make sure the beads remain in solution for the next step.
3. Perform A-tailing. Using the HTP Library Preparation (PCR-free) Kit, prepare the A-tailing Master Mix.
   1. Mix 5 μL of 10x A-tailing buffer (Final Concentration 1x) and 3 μL of A-tailing enzyme (total final volume of 8 μL of A-tailing Master Mix).
   2. Pipette 8 μL of the A-tailing Master Mix into each DNA specimen containing beads from step 7.2.4, mixing 8 μL of A-tailing Master Mix to 42 μL of End-Repaired DNA containing beads (total final volume of 50 μL). Place in a thermocycler at 30 °C for 30 min. Hold at 4 °C indefinitely.
   3. Pipette 1.8 volumes, or 90 μL, of PEG/NaCl SPRI solution (a component of the HTP Library Preparation Kit (PCR-free; 96 reactions)) to the A-tailed DNA. Purify the A-tailed DNA according to steps 4.3-4.6. Elute the A-tailed DNA in 30 μL of TE, pH 8.0. Keep the beads in solution for the next step.
4. Perform Adapter ligation. Using the HTP Library Preparation (PCR-free) Kit, prepare the Adapter Ligation Master Mix.
   1. Mix 10 μL of 5x ligation buffer (Final Concentration 1x), 5 μL of DNA Ligase, and 5 μL of annealed hairpin adapter (40 μM) from step 7.1.3. Ensure the Final Concentration 4 μM for a total Adapter Ligation Master Mix of 20 μL.
   2. Pipette 20 μL of Adapter Ligation Master Mix into each eluted DNA specimen containing beads from step 7.3.3 (total final volume of 50 μL per sample).
   3. Place in a thermocycler at 20 °C for 1 h. Hold at 4 °C indefinitely.
   4. Transfer 1x volumes, or 50 μL, of PEG/NaCl SPRI solution to the adapter-ligated DNA and purify according to steps 4.3-4.6. Elute with 30 μL of TE, pH 8.0, and decant the supernatants into a new semi-skirted PCR plate. Combine and quantify the DNA using the dsDNA BR assay. If needed, store purified adapter-ligated DNA for up to 1 month at -20 °C.
5. Prepare the Lambda Exonuclease/Exonuclease I (E. coli) Master Mix (functions in eliminating single- or double-stranded DNA without adapters ligated to both ends).
   1. Take 1 μg of Adapter Ligated DNA from step 7.4.4, diluting it to 40 μL. Mix 5 μL of 10x Exonuclease I reaction buffer (Final Concentration 1x), 4 μL of Lambda Exonuclease (Final Concentration 0.4 U/μL), and 1 μL of Exonuclease I (E. coli) (Final Concentration 0.4 U/μL) for a total volume of the Lambda Exonuclease/Exonuclease I Master Mix of 10 μL.
   2. Pipette 10 μL of the Lambda Exonuclease/Exonuclease I Master Mix to 40 μL (1 μg) of adapter-ligated DNA (total volume of 50 μL). Place in a thermocycler at 37 °C for 1 h, then 75 °C for 10 min. Hold at 4 °C indefinitely.
   3. Pipette 1.8x volumes, or 90 μL, of XP beads to the Lambda Exonuclease/Exonuclease I-treated DNA. Purify according to instructions given in steps 4.3-4.6. Elute in 40 μL of TE, pH 8.0. Make sure the beads remain in solution for the next enzymatic step.
6. Treat with USER Enzyme and T4 Polynucleotide Kinase (PNK). Prepare the USER/T4 PNK Master Mix (required to release the 4-bp overhangs and to prepare the ligation-ready DNA ends necessary for the subsequent ligation reaction).
   1. Mix 5 μL of 10x T4 DNA Ligase Buffer (Final Concentration 1x), 3 μL of USER Enzyme (Final Concentration 0.05 U/μL), and 2 μL of T4 PNK (Final Concentration 0.4 U/μL), for a total volume of the USER/PNK master mix of 10 μL.
   2. Pipette 10 μL of the USER Enzyme/T4 PNK Master Mix to 40 μL of lambda and exonuclease I-treated DNA specimen containing beads from step 7.5.3 for a total volume of 50 μL. Place in a thermocycler at 37 °C for 1 h. Hold at 4 °C indefinitely.
   3. Pipette 1.8x volumes, or 90 μL, of PEG/NaCl SPRI solution to the DNA treated with USER/T4 PNK and purify according to steps 4.3-4.6. Elute in 35 μL of TE, pH 8.0. Decant the supernatant into a new semi-skirted PCR plate. Combine and quantify the DNA using the dsDNA HS assay.
7. Perform Intramolecular circularization. Prepare the Circularization Master Mix.
   1. Mix 8 μL of Nuclease-free H₂O, 10 μL of 10x T4 DNA ligase buffer (Final Concentration 1x), and 2 μL of T4 DNA ligase (Final Concentration 8 U/μL), for a total volume of the Circularization Master Mix of 20 μL.
   2. Pipette 20 μL of the Circularization Master Mix to 500 ng of DNA treated with USER/PNK from step 7.6.3. Dilute 500 ng of DNA treated with USER/PNK in 80 μL, then add 20 μL of the Circularization Master Mix (total volume of 100 μL). Incubate in a thermocycler at 16 °C for 16 h (overnight).
   3. Add 1x volumes, or 100 μL, of XP beads to the circularized DNA and purify according to steps 4.3-4.6. Elute in 38 μL of TE, pH 8.0. Decant the supernatant into a new semi-skirted PCR plate.
8. Treat with Plasmid-Safe ATP-Dependent DNase. Prepare the Plasmid-Safe ATP-Dependent DNase Master Mix (Required for the degradation of residual linear DNA).
   1. Mix 5 μL of 10x Plasmid-safe Reaction Buffer (Final Concentration 1x), 2 μL of ATP (Final Concentration 1 mM), and 5 μL of Plasmid-Safe ATP-Dependent DNase (Final Concentration 1 U/μL), for a total volume of the Plasmid-Safe Master Mix of 12 μL.
   2. Pipette 12 μL of the ATP-dependent DNase Master Mix to 38 μL of circularized DNA from step 7.7.3 (total volume of 50 μL). Incubate in a thermocycler at 37 °C for 1 h, then 70 °C for 30 min. Hold at 4 °C indefinitely.
   3. Pipette 1x volumes, or 50 μL, of XP beads into the DNA treated with Plasmid-Safe ATP-Dependent DNase and purify according to steps 4.3-4.6. Elute in 15 μL of TE, pH 8.0. Decant the supernatant into a new semi-skirted PCR plate.
   4. Combine the DNA and quantify using the dsDNA HS assay. If needed, store the circularized DNA at -20 °C for up to several months.

8. Cleaving enzymatically purified, circularized gDNA in vitro (2 h)

1. Perform in vitro cleavage with the Cas9:gRNA complex. Prepare the in vitro Cleavage Master Mix. Mix 5 μL of 10x Cas9 Buffer (Final Concentration 1x), 4.5 μL of S. Pyogenes Cas9 (Final Concentration 90 nM), and 1.5 μL of gRNA (Final Concentration 90 nM) for a total Cleavage Master Mix volume of 11 μL.
2. Keep the cleavage master mix at RT for 10 min to form Cas9:gRNA RNP complexes.
3. Dilute 125 ng of Plasmid-Safe DNase-treated DNA from step 7.8.3 to a final volume of 39 μL. Then, add 11 μL of the Cleavage Master Mix to 39 μL Plasmid-Safe DNase-treated DNA for a total volume of 50 μL.
   NOTE: Include a negative control specimen in this step, which comprises circularized DNA mixed with Cas9 buffer, without the Cas9:gRNA complex.
4. Incubate in a thermocycler for 1 h at 37 °C. Maintain at 4 °C indefinitely. Add 50 µL (1x volume) of XP beads to the in vitro-cleaved DNA and purify the DNA following steps 4.3-4.6. Elute in 42 µL of TE buffer, pH 8.0. Ensure the beads stay in solution for the next step.

9. Preparation of the next-generation sequencing library (4 - 6 h)

1. Perform A-tailing. Prepare the A-tailing Master Mix.
   1. Mix 5 μL of 10x A-tailing Buffer (Final Concentration 1x), and 3 μL of A-tailing Enzyme (total volume of the A-tailing master mix is 8 μL).
   2. Pipette 8 μL of the A-tailing Master Mix to 42 μL of eluted DNA sample containing beads from step 8.4 (total volume of 50 μL). Place in a thermocycler for 30 min at 30 °C. Hold at 4 °C indefinitely.
   3. Pipette 1.8x volumes of PEG/NaCl SPRI solution, or 90 μL, to the A-tailed DNA and purify the DNA according to steps 4.3-4.6. Elute in 25 μL of TE, pH 8.0. Make sure to keep the beads in solution for the following step.
2. Perform Adapter ligation. Prepare the Adapter Ligation Master Mix.
   ​NOTE: Single-use aliquots of NEB adapters must be prepared in order to prevent the formation of adapter dimers caused by freeze-thaw-induced hydrolysis of the 3' T'.
   1. Mix 10 μL of 5x Ligation Buffer (Final Concentration 1x), 5 μL of DNA Ligase, and 10 μL of Adapter for sequencing (Final concentration is 3 μM), for a total of 25 μL.
   2. Pipette 25μL of the Adapter Ligation Master Mix into 25 μL of A-tailed DNA sample containing beads from step 9.1.3. Place in a thermocycler for 1 h at 20 °C. Hold at 4 °C indefinitely.
   3. Pipette 1x volumes, or 50 μL, of PEG/NaCl SPRI solution to the adapter-ligated DNA and purify the DNA according to steps 4.3-4.6. Elute in 47 μL of TE, pH 8.0. Make sure the beads remain in solution for the next enzymatic step.
3. Perform treatment with USER Enzyme (generates a single nucleotide gap at uracil residues).
   1. Add 3 μL of USER enzyme (included in Dual Index Primers Kit) to the adapter-ligated DNA specimen containing beads from step 9.2.3. Incubate at 37 °C for 15 min.
   2. Add 35 µL (0.7x volume) of PEG/NaCl SPRI solution to the USER enzyme-treated DNA and purify according to steps 4.3-4.6. Elute in 20 µL of TE buffer, pH 8.0. Transfer the supernatant to a new semi-skirted PCR plate and measure the DNA concentration using the dsDNA HS assay. The expected concentration should be approx. 2-5 ng/µL.
4. (Optional) Before proceeding to the next step, DNA size selection can be carried out using PippinHT. Use the 1.5% PippinHT cassette with a size range of 250-850 bp. The resulting samples can be directly utilized in the PCR in the next step.
5. Perform PCR for barcode addition
   NOTE: Ensure that the chosen primer sequence combinations for each sample are unique. If possible, each sample should have unique i5 and i7 barcodes.
   1. Prepare the PCR Master Mix to add paired-index barcodes. Mix 5 μL of Nuclease-free H₂O, 25 μL of 2x HotStart Ready Mix (Final Concentration 1x), 5 μL of i5 Primer (Final Concentration 1 μM), and 5 μL of i7 Primer (Final Concentration 1 μM) (total master mix volume of 40 μL).
   2. Pipette 40 μL of the PCR Master Mix to 10 μL of purified DNA treated with USER enzyme (approx. 20 ng) from step 9.3.2 (total volume of 50 μL).
   3. Choose the following PCR thermocycling conditions: Denaturation: 98 °C for 45 s for 1 cycle, Denaturation: 98 °C for 15 s for 20 cycles, Annealing: 65 °C for 30 s for 20 cycles, Extension: 72 °C for 30 s for 20 cycles, Final Extension: 72 °C for 1 min for 1 cycle, Hold: 4 °C indefinitely.
   4. Add 0.7x volumes, or 35 μL, of XP beads to the PCR product and purify according to steps 4.3-4.6. Elute in 30 μL of TE, pH 8.0. Decant the supernatant into a new semi-skirted PCR plate. If needed, store the circularized DNA at -20°C for up to several months.
      NOTE: Run a sample of the PCR on Tapestation to control for the quality of the library and to assess the formation of adapter dimers. If adapter dimers are detected, repeat step 9.5.4.

10. Quantification of CIRCLE-seq libraries by droplet digital PCR (dd_PCR) (6 h)

NOTE: Quantification may also be performed using qPCR, Tapestation, or a similar method.

1. Start with 5 μL of DNA from the library (PCR step 9.5.4) mixed well with 45 μL of nuclease-free TE, then make 1:10 serial dilutions of each sample in 50 μL volumes, ranging from 10^-1 to 10^-8 dilution.
   1. Set up the dd_PCR Master Mix stock solution. Mix 11 μL of 2x dd_PCR mix for probes (Final Concentration 1x), 0.055 μL of Probe oSQT1310 (Final Concentration 250 nM), 0.055 μL of Probe oSQT1311 (Final Concentration 250 nM), 0.099 μL of Primer oSQT1274 (Final Concentration 450 nM), 0.099 μL of Primer oSQT1275 (Final Concentration 450 nM), and 6.292 μL of Nuclease-free H₂O, for a total dd_PCR Master Mix volume of 17.6 μL. Prepare a master mix for all samples to ensure the volumes are sufficient for accurate pipetting.
   2. Assay the three lowest dilutions (10^-6,10^-7 and 10^-8) in duplicate (in a 96-well plate). A non-template control (NTC) must be used.
   3. Pipette 17.6 μL of dd_PCR Master Mix to each specimen as follows: Mix 17.6 μL with 4.4 μL of the sample (adding nuclease-free H₂O to the well with NTC), for a total volume of 22 μL. Seal the plate, then centrifuge at 2000 x g for 1 min at RT.
2. Perform Droplet generation, thermocycling, and analysis. Using a Droplet Reader PCR system, transfer a DG8 cartridge (8-well) into the cartridge holder. In the oil row of the cartridge, dispense 70 μL of droplet generation oil for probes.
   ​NOTE: Take 20 μL of the sample from step 10.1.2. and add it to the sample row of the 8-well cartridge, covering the cartridge with the DG8 rubber gasket, placing it into the droplet generator, and then closing it to begin the process (automatic). When complete, remove the cartridge and move 40 μL from the droplet row of the 8-well cartridge into a semi-skirted 96-well PCR plate, making sure to pipette slowly.
   1. Place the heat block into the PX1 PCR plate sealer. The sealer will begin heating up to 180 °C when turned on. Put the foil heat seal on the plate, ensuring the red line is on top. Place the plate into the PX1 and press Seal.
   2. Choose the following Thermocycler conditions: Enzyme Activation: 95 °C for 10 min for 1 cycle, Denaturation: 94 °C for 30 s for 40 cycles, Annealing/Extension: 60 °C for 1 min for 40 cycles, Enzyme Deactivation: 98 °C for 10 min for 1 cycle, Hold: 4 °C indefinitely.
   3. On the droplet reader, open the compatible software and select which wells are to be read. Select ABS as the experiment type and dd_PCR Supermix for probes. Choose Ch1 Unknown for Target 1 and Ch2 Unknown for Target 2. Select Apply, then OK. Put the plate into the droplet reader. For the dye set, choose FAM/HEX and click on Run.
3. Analyze the dd_PCR results. Gate the double positive droplet population using the negative control as a reference. Calculate the average of the duplicate values, then multiply by the dilution factor and the 5-fold dilution factor of the dd_PCR.
   NOTE: The total copies per microliter are calculated as follows: total copies per microliter = mean value × 5 × dilution factor, where 'mean value' represents the average quantification value from Ch1 and Ch2.
4. Combine all samples into a single library at equimolar concentrations. The 1x pooled library should contain approx. 4.5 x 10⁹ molecules and have a total volume of 5 μL.

11. Next-generation sequencing

1. Submit the samples for sequencing to an external agency, ensuring the correct adapter sequences are noted.

12. CIRCLE-seq data analysis (1 - 3 h)

1. Install Python version 2.7, Burrows-Wheeler Aligner (BWA), and SAMtools. Download the reference genome (e.g., hg38) from http://hgdownload.cse.ucsc.edu/goldenPath/hg38/bigZips/hg38.fa.gz.
   NOTE: If the genome of the target species is unavailable, the CIRCLE-seq computational workflow can be run in a reference-independent mode. In this situation, this step can be skipped.
2. Download and install the CIRCLE-seq pipeline with the following commands: (1) git clone https://github.com/tsailabSJ/circleseq.git, (2) cd circleseq, (3) pip install -r requirements.txt.
3. Create a manifest file in YAML format (.yaml). Below is a sample manifest that can be used with the example dataset provided in the CIRCLE-seq software to test the workflow.
   NOTE: (1) reference genome: data/input/CIRCLEseq_test_genome.fa; (2) analysis_folder: data/output; (3) bwa: bwa; (4) samtools: samtools; (5) read_threshold: <Value>; (6) window_size: <Value>; (7) mapq_threshold: <Value>; (8) start_threshold: <Value>; (9) gap_threshold: <Value>; (10) mismatch_threshold: <Value>; (11) merged_analysis: True; (12) samples: U2OS_EMX1; (13) target: GAGTCCGAGCAGAAGAAGAANGG; (14) read1: data/input/EMX1.r1.fastq.gz; (15) read2: data/input/EMX1.r2.fastq.gz; (16) controlread1: data/input/EMX1_control.r1.fastq.gz; (17) controlread2: data/input/EMX1_control.r2.fastq.gz; (18) description: U2OS. The following manifest values were used: read_threshold: 4, window_size: 3, mapq_threshold: 50, start_threshold: 1, gap_threshold: 3, mismatch_threshold: 6
4. Define the reference genome FASTA file, the output directory for analysis, and the paths to the BWA and SAMtools commands. Define the target sequences and the paths to the demultiplexed FASTQ files for both the nuclease-cleaved and control samples. Multiple experiments can be processed simultaneously in batch mode by including all of them in a single manifest file.
5. Execute the following command for standard reference-based analyses: (1) python /path/to/circleseq.py all - manifest; (2) /path/to/manifest.yaml.
6. Alternatively, execute the following command for standard non-reference-based analyses: (1) Python /path/to/circleseq.py reference-free - manifest; (2) /path/to/manifest.yaml.
7. When executing the full pipeline, find the output results of each step in a distinct output_folder designated for that specific step.

Résultats

Here, CIRCLE-seq is utilized to investigate the nuclease-induced cleavage sites of Cas9 in a complex with the gRNA designed to target the adeno-associated virus integration site 1 (AAVS1) using DNA isolated from induced pluripotent stem cells (iPSCs). This gRNA was previously described in our publication²⁷. Approximately 25 μg of gDNA was isolated from iPSCs, sheared through focused ultrasonication, and size selected using AMPure XP bead purification to yield fragments of approx. 300 bps. From this 25 μg of DNA, approx. 2-5 ng of DNA was successfully circularized for in vitro Cas9:gRNA cleavage. The entire procedure is depicted in Figure 1.

Following a CIRCLE-seq procedure and analysis using our computation workflow, a visualization of all detected on- and off-target cleavage sites is presented in Figure 2A. The CIRCLE-seq pipeline also provided 'merged reads', analyzed via R statistical software to yield a Manhattan plot showing the detected nuclease-induced cleavage sites mapped along each chromosome (Figure 2B).

Figure 1: CIRCLE-seq workflow schematics. Major steps of the protocol are indicated. Please click here to view a larger version of this figure.

Figure 2: CIRCLE-seq visualization and Manhattan Plot. (A) Alignment of off-target sites against the intended target for the AAVS1 locus. The target sequence is displayed at the top, where off-targets are ranked by read count in descending order. Differences in the original target sequences are shown by colored nucleotides. A sample of the top unintended cleavage sites for the AAVS1 locus is shown. (B) Manhattan plot illustrating the detected unintended cleavage sites for the AAVS1 locus. Bar heights represent read counts for each chromosomal position. Please click here to view a larger version of this figure.

Primer	Sequence (5'-3')	Comments/Description
AAVS1 Single guide RNA (sgRNA)	GGGGCCACUAGGGACAGGAU	For AAVS1 Locus' Fluorescent Protein Knock-In
AAVS1 Forward Primer	GCTCTGGGCGGAGGAATATG	For gRNA in vitro cleavage test
AAVS1 Reverse Primer	ATTCCCAGGGCCGGTTAATG	For gRNA in vitro cleavage test
oSQT1288	/5Phos/CGGTGGACCGATGATC /ideoxyU/ATCGGTCCACCGaT	CIRCLE-seq hairpin adapter
oSQT1274	AATGATACGGCGACCACCGAG	TruSeq F1
oSQT1275	CAAGCAGAAGACGGCATACGAGAT	TruSeqF2
oSQT1310	/56-FAM/CCTACACGA/ZEN/CGCTCTTCCGATCT/3IABkFQ/	TruSeq probe
oSQT1311	/5HEX/TCGGAAGAG/ZEN/CACACGTCTGAACT/3IABkFQ/	TruSeq probe

Table 1: Sequences of gRNA and primers used for CIRCLE-seq analysis of AAVS1 locus.

Discussion

Here, CIRCLE-seq is demonstrated to be an unbiased and highly sensitive technique for identifying nuclease-induced DSBs across the genome resulting from targeting the AAVS1 locus in gDNA derived from iPSCs. The AAVS1 site within iPSCs is well-known as a safe harbor locus that is often used as an integration site of exogenous genes using CRISPR-Cas9²⁸. Our recent report studied the potential of EGFP-labeled iPSCs by performing CRISPR-mediated integration of a constitutively expressed EGFP reporter into the AAVS1 site, which enables labeling and tracking of both iPSCs and differentiated iPSCs due to the persistence of EGFP throughout the cell's lineage²⁷. This iPSC line can be used in vivo to evaluate the organismal distribution of iPSC-derived cells after transplantation. As this cell line has been CRISPR-modified and is also being used to test the clinical application of iPSCs, it is imperative that the potential AAVS1 off-target sites are known and interrogated to ensure safety and efficacy, making it an ideal locus to test CIRCLE-seq.

A notable difference between Butterfield et al., previously published study²⁷ and this one was the use of a modified gRNA to target the AAVS1 locus. A gRNA can be designed to improve genome editing accuracy²⁴. The guide sequence is the most critical factor that influences on- and off-target efficiencies. Therefore, the chosen gRNA was tested against several other guides and found to have superior fidelity. In addition, a methods article on reprogramming human fibroblasts substantiated the finding that synthetic capped mRNAs containing modified nucleobases benefit from low activation of antiviral responses^29,30. While low immunogenicity might not be relevant in an in vitro assay, it becomes crucial when the ultimate goal is to develop a clinically relevant therapy that can be used in living cells.

CIRCLE-seq holds many advantages over similar methods. For instance, Digenome-seq sequences both nuclease-cleaved and uncleaved gDNA, using ~400 million reads²¹. This results in a high background, making it challenging to filter out low-frequency bona fide cut sites. CIRCLE-seq only uses ~3-5 million reads due to the enrichment of nuclease-cleaved gDNA, resulting in low background. Additionally, Digenome-seq and a similar method, SITE-seq, rely on sequencing a single nuclease-cleaved DNA end. In contrast, CIRCLE-seq reads include both ends of the cut site, enabling the identification of off-target sites without the need for a reference^21,22,26.

An advantage of CIRCLE-seq is its greater sensitivity compared to the methods that rely on cell culture, such as GUIDE-seq. When the two methods were compared, CIRCLE-seq was able to capture all off-target sites detected by GUIDE-seq and uncovered additional unintended cleavage sites that GUIDE-seq had missed. However, a notable difference is that GUIDE-seq may be hindered by the epigenetic landscape, whereas CIRCLE-seq can access the entire genome.

As an in vitro assay, CIRCLE-seq presents several limitations, the first of which is the detection of false positives. Whereas epigenetics will impede nuclease activity at certain sites in vivo, ultrasonication removes these obstacles in vitro, allowing off-target activity at locations that may not normally be accessible in a cellular context. Furthermore, Cas9 is present at high concentrations in this in vitro assay, permitting cleavage that would not otherwise be possible in vivo. This assay also requires a relatively large amount of starting gDNA, which, depending on available resources, can negate the use of this protocol. Lastly, it is possible for some off-target sites to be undetectable due to the limitations of current next-gen sequencing technologies.

A recent study used an in silico approach whose algorithm identified a number of relevant parameters with which to compare different nuclease characterization methods, including CIRCLE-seq and GUIDE-seq³¹. Two of the relevant parameters were 'cut-site enrichment' and '% false positives'. Interestingly, CIRCLE-seq's false-positive rate was calculated at 88%, but its cut-site enrichment was folds higher than the other in vitro methods. Comparative analyses of every method revealed that GUIDE-seq was the best performer, as it demonstrated the greatest on-target specificity with only a moderate false-positive rate³². This does not invalidate CIRCLE-seq, but rather hints at the possibility of utilizing CIRCLE-seq and GUIDE-seq in tandem, validating CIRCLE-seq's findings with GUIDE-seq since the former has higher sensitivity, while the latter is a cell-based method with high cut-site enrichment. Data also indicate that amplicon-based next-generation sequencing (NGS) should be the preferred method for identifying genuine off-target modifications at potential candidate sites³¹. This data suggests a potential strategy of using CIRCLE-seq, followed by GUIDE-seq, and then amplicon-based NGS to examine off-target effects.

matériels

Name	Company	Catalog Number	Comments
0.2-mL Thin-walled Tubes and Flat Caps	ThermoFisher Scientific	AB1114
1.5% PippinHT cassette	Sage Science	HTC1510
10 mL Serological Pipettes	FisherScientific	12567603
15 mL Conical Tube	FisherScientific	339651
150 x 25 mm Tissue Culture Dish	FisherScientific	877224
25 mL Reagent Reservoir	FisherScientific	2138127C
2x Kapa KiFi HotStart Ready Mix	Kapa Biosystems	KK2602
5 mL Serological Pipettes	FisherScientific	170355
50 mL Conical Tube	FisherScientific	339653
50x TAE Electrophoresis Buffer	ThermoFisher Scientific	B49
6 Well Cell Culture Plate	Corning	3516
Agencourt AMPure XP Reagent, 60 mL	Beckman Coulter	A63881
Benchtop Microcentrifuge	Eppendorf	5400002
BsaI-HF	New England BioLabs
Buffer QX1	Qiagen	20912
Cas9 nuclease, Streptococcus Pyogenes	New England BioLabs	M0386M
CIRCLE-seq Library Preparation and NGS
Corning Matrigel hESC-Qualified Matrix	Corning	354277	Extracellular matrix (ECM) for culturing iPSCs
ddPCR SuperMix for Probes	Bio-Rad	1863010
DG8 Cartridge Holder	Bio-Rad	1863051
DG8 Cartridges	Bio-Rad	1864008
DG8 Gaskets	Bio-Rad	1863009
DPBS, no calcium, no magnesium	ThermoFisher Scientific	14190144	Made by Invitrogen
Droplet Generation Oil for Probes	Bio-Rad	1863005
Droplet Reader Oil	Bio-Rad	1863004
EDTA (0.5 M)	ThermoFisher Scientific	15575020
EDTA (0.5 M), pH 8.0, RNase-free	ThermoFisher Scientific	AM9260G	Made by Invitrogen
Eppendorf  ThermoMixer	Eppendorf	5384000020
Equipment
Ethyl Alcohol, Pure	Sigma-Aldrich	E7023
Exonuclease I	New England BioLabs	M0293L	E. coli
Filter Unit	FisherScientific	FB0875713
Filtered Sterile Pipette Tips
Focused Ultrasonicator	Covaris	ME220
Genomic DNA Isolation
Genomic DNA Shearing
Gentra Puregene Cell Core Kit	Qiagen	158043
gRNAs	Synthego
HCl	ThermoFisher Scientific	A144500
Heracell VIOS Tri-gas Humidified Tissue Culture Incubator	ThermoFisher Scientific	51030411	Need for culturing and expanding iPSCs (37 °C/5% CO2/5% O2)
High Sensitivity D1000 DNA ScreenTape	Agilent	5067-5584
High Sensitivity D1000 Reagents	Agilent	5067-5585
HTP Library Preparation Kit	Kapa Biosystems	KK8235
HyClone Antibiotic Antimycotic Solution	Cytiva	SV30079.01
IDTE pH 8.0 (1x TE Solution)	Integrated DNA Technologies	11050204
Inverted Microscope			Need for imaging iPS colonies in bright-field and fluorescent channels
iPSC Culture
Isopropanol	Sigma-Aldrich	190764
Lambda Exonuclease	New England BioLabs	M0262L
Loading Tips, 10 pack	Agilent	5067-5599
Magnum FLX Enhanced Universal Magnet Plate	Alpaqua	A00400
Microcentrifuge Tube	Axygen	31104051
Microtube AFA Fiber Pre-Slit Snap-Cap	Covaris	520045
mTeSR-1 5x Supplement	StemCell Technology	85852
mTeSR-1 Basal Medium (400 mL)	StemCell Technology	85851	Media for maintaining iPSC in culture
Nanodrop 8000 Spectrophotometer	ThermoFisher Scientific	ND-8000-GL
NEBNext Multiplex Oligos for Illumina	New England BioLabs	E7600S	Dual Index Primers Set 1
Optical tube strip caps, 8x strip	Agilent	401425
Optical tube strips, 8x strip	Agilent	401428
Other Reagents
PCR Plate Sealer	Bio-Rad	PX1	Model Number PX1
PEG/NaCl SPRI Solution	Kapa Biosystems
Phusion Hot Start Flex 2x Master Mix	New England BioLabs	M0536L
Pierceable Foil Heat Seal	Bio-Rad	1814040
Plasmid-Safe ATP-dependent DNase	Epicentre	E3110K
Primers, Adapters and Probes	IDT		Sequences are listed in Table 1
Proteinase K	Qiagen	19131
QIAquick Gel Extraction Kit	Qiagen	28704
QIAxcel Gel Analysis System	Qiagen	9001941
Qubit Assay Tubes	ThermoFisher Scientific	Q32856
Qubit dsDNA BR Assay Kit	ThermoFisher Scientific	Q32853
Qubit dsDNA BR Assay Kit	ThermoFisher Scientific	Q32853
Qubit dsDNA HS Assay Kit	ThermoFisher Scientific	Q32854
Qubit Fluorometer	ThermoFisher Scientific	Q33226
QX200 Droplet Digital PCR System	Bio-Rad	1864001	Contain a QX200 droplet generator and a QX200 droplet reader
RNase A	Qiagen	19101
Semi-skirted PCR Plate	ThermoFisher Scientific	14230244
SeqPlaque GTG Agarose	Lonza	50110
SYBR Safe DNA Gel Stain	ThermoFisher Scientific	S33102
T4 DNA Ligase	New England BioLabs	M0202L
T4 Polynucleotide Kinase (PNK)	New England BioLabs	M0201L
T-75 Flasks	FisherScientific	7202000
Tapestation 4150	Agilent	G2992AA
Thermocycler with programmable temperature-stepping functionality	Bio-Rad C1000 Touch
Tris base	ThermoFisher Scientific	BP1521
Twin.tec PCR Plate	Eppendorf	E951020346	96 wells, semi-skirted, green
UltraPure DNase/RNase-Free Distilled Water	ThermoFisher Scientific	10977015
USER Enzyme	New England BioLabs	M5505L