Targeted DNA ADP-ribosylation triggers templated repair in bacteria and base mutagenesis in eukaryotes

Polymerase-blocking assays

WT and inactivated (E170A) EPEC DarT2 proteins were expressed using the cell-free myTXTL master mix (Arbor Biosciences). Linear DarT expression templates were amplified from plasmids or ordered as synthetic gene fragments (Integrated DNA Technologies) and contained a T7 promoter and a T7 terminator (Supplementary Table 2). Cell-free expression was performed in 12-µl reactions, comprising 9 µl of myTXTL master mix, 4 nM of EPEC DarT2 template, 0.4 nM of a T7 RNA polymerase-encoding plasmid and 4 µM of the RecBCD inhibitor GamS to prevent degradation of the linear DNA templates. The reactions were incubated for 16 h at 29 °C.

For ADP-ribosylation of ssDNA templates, the ADP-ribosylation assay was adapted from prior work with slight alterations¹³. Briefly, 5 µl of the TXTL reaction mix was incubated with 10 µM of the ssDNA oligo, 50 µM NAD⁺, 50 mM Tris-HCl pH 8, 150 mM NaCl, 10 mM EDTA and sterile nuclease-free water to reach a final volume of 20 µl and incubated for 30 min at 30 °C. Afterward, the oligos were separated from the mix using the Oligo clean and concentrator kit (Zymo).

To assess whether DNA ADP-ribosylation blocks DNA polymerases in vitro, the DarT-treated oligos were first annealed to the 5′ 6-Fam-tagged primer CKo20 at a final concentration of 10 µM in 1× NEBuffer 2 by heating the mixture to 94 °C and gradually cooling it to room temperature. Next, 2 µl of the annealed product was mixed with 0.5 U of Klenow fragment (New England Biolabs), 33 µM dNTPs and 1× NEBuffer 2 in a total volume of 12.5 µl and incubated for 15 min at 37 °C. To stop the reaction, EDTA was added to a final concentration of 10 mM and the samples were incubated at 75 °C for 20 min.

To visualize the block of polymerization, 4 µl of the polymerization product was mixed with 4 µl of loading dye (containing 95% formamide, 0.03% SDS, 18 mM EDTA, 23 µM xylene cyanol and 19 µM bromophenol blue) and loaded onto a preheated denaturing polyacrylamide gel (8 M urea and 20% polyacrylamide (19:1)). The gel was run at 250 V for 30 min and visualized under ultraviolet light before and after staining with SYBR gold (Thermo Fisher).

Microbial strains, handling and growth conditions

All bacterial and yeast strains used in this study are listed in Supplementary Table 2. Unless otherwise specified, E. coli TOP10 was used for plasmid cloning and propagation and was grown at 37 °C in Luria–Bertani (LB) liquid medium (10 g L⁻¹ tryptone, 5 g L⁻¹ yeast extract and 10 g L⁻¹ NaCl) shaking orbitally at 200 rpm or on LB solid medium (15 g L⁻¹ agar) at 37 °C, containing kanamycin (50 mg L⁻¹), carbenicillin (100 mg L⁻¹) or chloramphenicol (34 mg L⁻¹), when appropriate. The E. coli kanR* strain (CBS-4802) began as strain CB330 (E. coli MG1655 P_J23110–araFGH ∆araBAD), selected for uniform arabinose induction, to which two chromosomal modifications were made. First, the ∆lacZ phenotype (W519*) was generated by CBE-mediated deamination of 5′-ACC-3′ to 5′-ATT-3′ (positions 364,749 and 364,750 in MG1655), resulting in a premature stop codon; this edit was not used in this work. Second, a defective kanR expression construct (kanR*) (annotated sequence of the genomic locus in Supplementary Table 2) containing a premature stop codon (Q177*) and DarT2 motif 5′-TTTC-3′ was inserted between genes ybjM and grxA (positions 890,463–890,480 in MG1655) by Red-mediated recombination with Cas9 counterselection^1,36,65. The resulting E. coli MG1655 kanR* strain was used for all assays related to the kanR* gene. The kanR* strain was further used to generate ΔrecA, ΔrecB, ΔrecF, ΔrecT, ΔrecJ, ΔrecO, ΔxthA, ΔmutS and ΔuvrA mutants by Red-mediated recombination⁶⁶. Briefly, transformants of the E. coli kanR* strain carrying pKD46 (encoding λ Red-γ, Red-β and Red-exo) were cultured in l-arabinose at 30 °C until an optical density at 600 nm (OD₆₀₀) of ~0.6, made electrocompetent as previously described⁶⁶ and then transformed with a linear dsDNA template containing 40-nt homology arms to mediate deletion of the target gene. Next, pKD46 was cured from the bacteria by growing them at 37 °C, after which the bacteria were made electrocompetent and transformed with pCP20 and then grown at 42 °C to simultaneously express FLP recombinase and eliminate pCP20. Colonies were then screened for gene deletion by colony PCR and Sanger sequencing. For the substitution assays targeting the aaaD, punR, ygcQ and yheO genes, the E. coli MG1655 strain was used.

Salmonella enterica subsp. enterica serovar Typhimurium strain LT2 was used for all ADPr-TAE assays in Salmonella and was regularly grown at 37 °C in LB liquid medium shaking orbitally at 200 rpm or on solid LB medium. Carbenicillin (100 mg L⁻¹) and chloramphenicol (34 mg L⁻¹) were supplemented in the growth medium when necessary.

The S. cerevisiae BY4741 (Δtrp1, Δleu2) strain was used for all yeast experiments. Unless otherwise specified, S. cerevisiae was grown in nonselective liquid YPD medium (20 g L⁻¹ peptone, 10 g L⁻¹ yeast extract and 2% (w/v) d(+)-glucose) or on solid nonselective YPD medium (20 g L⁻¹ agar). To select for transformants, S. cerevisiae cells were grown on solid synthetic defined (SD) medium without tryptophan and leucine, containing 6.9 g L⁻¹ yeast nitrogen base without amino acids (Formedium, CYN0402), 0.64 g L⁻¹ complete supplement mixture without tryptophan and leucine (Formedium, DCS0569), 20 g L⁻¹ d(+)-galactose (Sigma-Aldrich, 15522-250G-R) and 20 g L⁻¹ agar (Th. Geyer, 214510).

Plasmid construction

Annotated sequences of all plasmids used in this study are provided in Supplementary Table 2. Unless otherwise specified, general cloning methods such as KLD (KLD enzyme mix, M0554S) or Gibson assembly (NEBuilder HiFi DNA assembly master mix, E2621X) were used to assemble linear dsDNA fragments into plasmids. Linear dsDNA fragments were amplified with Q5 high-fidelity 2× master mix (New England Biolabs, M0492L) and purified using the NucleoSpin gel and PCR cleanup kit (Macherey-Nagel, 740609.50). Plasmid sequences were verified by full plasmid sequencing (Plasmidsaurus) or Sanger sequencing (Microsynth Seqlab).

To generate the append editors expressed in plants, the codon-optimized DNA sequence for DarT2^D was commercially synthesized (Twist Bioscience) with a previously reported N7-NLS for expression in N. benthamiana⁶⁷, while the zCas9i (Zea mays codon-optimized Cas9 coding sequence with 13 introns) was obtained from Addgene (kit 1000000171)⁶⁸. Both fragments were amplified using the iProof high-fidelity PCR kit (Bio-Rad, 1725331). The dDarT, nzCas9i and dzCas9i variants were generated using inverse PCR. Three gRNAs targeting the phytoene desaturase 1 gene (PDS1) (Supplementary Table 2) were cloned by annealing complementary oligos into an AtU6 gRNA cassette. Gene fragments were assembled using the GoldenBraid cloning strategy⁶⁹.

kanR* reversion

To assess ADPr-TAE in E. coli, an overnight culture of strain CBS-4802 was backdiluted 100-fold, grown to an OD₆₀₀ of 0.6–0.8 and then rendered electrocompetent in 10% glycerol. For transformation, 40 μl of electrocompetent cells were mixed with 9 fmol of the relevant plasmid(s) and transferred to an ice cold 1-mm electroporation cuvette (Bio-Rad Laboratories, 1652089). Cells were electroporated using the GenePulser Xcell microbial system (Bio-Rad Laboratories, 1652662) and the following settings: 1.8 kV, 25 µF and 200 Ω. Next, cells were supplemented with 500 μl of SOC medium (5 g L⁻¹ yeast extract, 20 g L⁻¹ tryptone, 0.584 g L⁻¹ NaCl, 0.186 g L⁻¹ KCl, 2.4 g L⁻¹ MgSO₄ and 20 mM glucose) and recovered for 1 h at 37 °C, shaking orbitally at 200 rpm. Cells were collected by centrifugation at 3,000g, the supernatant was decanted and cells were resuspended in 2 ml of induction medium (LB, l-arabinose (0.2% w/v), carbenicillin (100 mg L⁻¹) and chloramphenicol (34 mg L⁻¹)) and incubated at 37 °C for 16 h, shaking orbitally at 200 rpm. Afterward, cell cultures were serially diluted in five tenfold steps in LB, from which 3 μl of each dilution was spotted on LB solid medium containing either carbenicillin and chloramphenicol to select for transformed cells or carbenicillin, chloramphenicol and kanamycin to select for transformed and edited cells. The spotted LB solid medium was then incubated for 16 h at 37 °C followed by counting colonies.

Replacement, deletion and insertion assays in E.
coli

For the E. coli replacement, deletion and insertion assays at the kanR* locus and the substitution assays at the aaaD, punR, ygcQ and yheO genes, an identical transformation and selection protocol was used as described above. However, after the 16-h incubation in the induction medium, 100 μl of the cell culture was plated on LB solid medium containing carbenicillin and chloramphenicol to obtain single colonies. Single colonies were resuspended in Q5 high-fidelity 2× master mix containing the appropriate primers and subjected to PCR amplification following the instructions of the manufacturer and extending the initial heating step of 98 °C to 5 min to mediate cell lysis and release of genomic DNA. Amplicons were purified and sequenced through Sanger sequencing.

Growth-based toxicity assay in E.
coli

The growth-based toxicity assay began by rendering strain CBS-5301 electrocompetent. Next, 9 fmol of plasmid CBS-4808 was transformed into strain CBS-5301 using the electroporation conditions described above. Transformants were recovered in 500 µl of SOC medium for 1 h at 37 °C, shaking orbitally at 200 rpm, then plated on LB solid medium supplemented with carbenicillin and incubated for 16 h at 37 °C. Next, a single colony was inoculated into 2 ml of LB medium containing carbenicillin, grown until an OD₆₀₀ of 0.6 and then made electrocompetent following the protocols described above. A second round of transformation was performed, using one of nine different editor plasmids (CBS-6738, CBS-6739, CBS-6741, CBS-6742, CBS-6743, CBS-6744, CBS-6745, CBS-4781 or CBS-4800), following the electroporation protocol described above. Transformed cells were allowed to recover in 500 µl of SOC medium for 1 h at 37 °C shaking orbitally at 200 rpm, plated on LB solid medium supplemented with carbenicillin, chloramphenicol and glucose (20 mM) and incubated for 16 h at 37 °C. Three individual colonies from each of the nine resulting strains (Supplementary Table 2) were then used to inoculate a 96-deep-well plate (Greiner Bio-One, 780271), containing 400 µl of LB medium supplemented with carbenicillin, chloramphenicol and glucose (20 mM) and covered with an adhesive gas-permeable membrane (Thermo Scientific, 241205). After incubating the deep-well plate for 16 h at 37 °C, the cell cultures were adjusted to an OD₆₀₀ of 0.1 using LB supplemented with carbenicillin, chloramphenicol and l-arabinose (0.2% w/v) in a new 96-well plate, reaching a final volume of 200 µl. The 96-well plate was then measured every 3 min over 12 h at 37 °C for absorbance at 600 nm on a BioTek Synergy Neo2 plate reader, shaking at 500 rpm.

Nonselective editing at kanR*

Transformations were performed as described above; however, after the 16-h incubation in induction medium, the cultures were centrifuged, the medium was discarded and genomic DNA was isolated using the Wizard gDNA purification kit (Promega, A1120). The kanR site was then amplified through PCR using the primer pair HBo-314 and HBo-315 and the Q5 high-fidelity 2× master mix for 25 cycles. Resulting amplicons were sequenced with Nanopore sequencing (Eurofins Genomics). For data analysis, FASTQ sequencing data files were aligned to a FASTA file of the unedited amplicon using MiniMap2 with option ‘map-ont’⁷⁰. SAMtools was used to convert the SAM files into BAM files, while concurrently sorting and indexing⁷¹. All further analysis was performed using R, after calling libraries tidyverse and GenomicAlignments⁷². A function was defined to take BAM files as an argument and then extract all alleles aligned to the 8-nt region of the templated edit as a list of characters. This function was applied to all BAM files to generate lists of alleles, which were tallied and compiled into a single data frame in long table format. Next, alleles were defined as unedited, edited or ambiguous and the fraction of each observation was computed. Samples were then grouped by editor and repair plasmids, after which the mean and s.d. were computed and then used to generate the bar plot. Further analysis was undertaken to search for base mutations at the ADPr site. The list of alleles in the initial data frame was filtered to retain only records containing a T-to-V mutation at the ADPr target position but otherwise matching the reference allele. Records were grouped by sample and SNVs were tallied, after which each was divided by the total number of observed alleles and multiplied by 100 to obtain the percentage of base mutations amongst all sequencing reads.

Whole-genome off-target assay in E.
coli

For identifying whole-genome off-target mutations, strain CBS-4802 was grown from a single colony in LB medium and made electrocompetent as described above. Electrocompetent CBS-4802 was then cotransformed with equimolar amounts (9 fmol) of CBS-6746 and one of several editor plasmids (CBS-3130, CBS-6738 or CBS-6740). Transformants were recovered in 500 μl of SOC for 1 h at 37 °C shaking orbitally at 200 rpm, after which the growth medium was replaced with 2 ml of LB, supplemented with carbenicillin, chloramphenicol and l-arabinose (0.2%), followed by incubation at 37 °C for 16 h shaking orbitally at 200 rpm. Next, the cultures were streaked onto LB solid medium supplemented with carbenicillin and chloramphenicol and incubated for 16 h at 37 °C to obtain individual colonies. Three colonies from each condition were placed in 2 ml of LB medium supplemented with carbenicillin and chloramphenicol and cultured for 16 h at 37 °C.

After incubation, cultures were centrifuged and the cell pellets were subjected to genomic DNA isolation using the Wizard genomic DNA purification kit. Isolated genomic DNA was fully sequenced using Nanopore sequencing (Plasmidsaurus). For data analysis, FASTQ sequencing data files were aligned to a FASTA file of E. coli MG1655 (GenBank: U00096.3) using Minimap2 with the ‘map-ont’ option⁷⁰. SAMtools was used to convert the SAM files into BAM files, while concurrently sorting and indexing⁷¹. Clair3 was run on the GalaxyEU server to call variants^73,74. Bcftools was used to query the VCF files for POS, REF, ALT, DP and AF fields and export the results into a CSV file⁷⁵. The sequencing depth at all positions in all BAM files was calculated by SAMtools and exported as a CSV file. All further analysis was performed in R after loading library tidyverse⁷². CSV files were loaded into a long-format data frame. This data frame was then filtered as follows: (1) SNVs were retained by filtering for records that contain only a single character in the REF and ALT fields; (2) SNVs already present in the parent strain were eliminated by filtering for records containing POS field values not found in parent strain POS field values; (3) SNVs mapped to regions known to have been modified during the creation of strain CBS-4802 were eliminated by filtering for records with POS field values not present in said regions; (4) records were filtered for AF field values greater than or equal to 0.25; (5) SNVs observed at a sequencing depth greater than or equal to the lowest quartile of all BAM files (Q1 ≥ 34) were retained; and (6) all SNVs were recoded to C > D and T > V, tallied and then used to generate a heat map.

Editing assays in S.
enterica

Electrocompetent S. enterica cells were transformed with 9 fmol of plasmid CBS-4800 and recovered in 500 μl of SOC medium following an identical protocol to that described above for E. coli. After recovery, the cells were collected through centrifugation at 3,000g, the supernatant was decanted and the cell pellet was resuspended in 100 μl of LB medium. The cell suspension was plated on LB solid medium containing chloramphenicol (34 mg L⁻¹) and incubated at 37 °C for 16 h. After incubation, a single colony was selected and used to create electrocompetent S. enterica cells harboring plasmid CBS-4800 following the protocol described above. Then, 22 fmol of the plasmids containing the repair template and the T sgRNA (Supplementary Table 2) were transformed in triplicate through electroporation into S. enterica cells harboring plasmid CBS-4800. The cells were recovered in 500 μl of SOC medium and collected through centrifugation at 3,000g, the supernatant was decanted and the cell pellet was resuspended in 2 ml of induction medium (LB, 0.2% (w/v) l-arabinose, 100 mg L⁻¹ carbenicillin and 34 mg L⁻¹ chloramphenicol) and grown at 37 °C for 16 h, shaking orbitally at 200 rpm. Next, 100 μl of the cell culture was plated on LB solid medium containing carbenicillin and chloramphenicol to obtain single colonies. Colonies were resuspended in Q5 high-fidelity 2× master mix containing the appropriate primers and subjected to PCR amplification following the instructions of the manufacturer and adding an initial heating step of 98 °C for 5 min to mediate cell lysis and release of genomic DNA. Amplicons were then purified using the NucleoSpin gel and PCR cleanup kit and sequenced through Sanger sequencing.

Templated editing assays in S.
cerevisiae

S. cerevisiae BY4741 (Δtrp1, Δleu2) cells were cotransformed with two plasmids, one bearing the specified editor variant and the other bearing a 6-bp substitution template flanked by 294-bp (upstream) and 232-bp (downstream) homology arms along with an FCY1 T sgRNA or NT sgRNA (Supplementary Table 2), following the lithium acetate method as previously described⁷⁶.

Briefly, single S. cerevisiae colonies were inoculated into 2 ml of liquid YPD medium (20 g L⁻¹ peptone, 10 g L⁻¹ yeast extract and 2% (w/v) d(+)-glucose) and grown for 16 h at 30 °C, shaking at 200 rpm on a rotary shaker. The cells were diluted to an OD₆₀₀ of 0.5 in 50 ml of YPD medium and cultured again at 30 °C, shaking at 200 rpm, until the cells reached an OD₆₀₀ of 2. The cells were then harvested by centrifugation at 3,000g for 5 min, the supernatant was decanted and the pellet was resuspended in 25 mL of sterile water. The centrifugation and resuspension step was repeated followed by another centrifugation at 3,000g for 5 min and resuspension in 1 ml of sterile water. The cell suspension was then centrifuged for 30 s at 13,000g, the supernatant was discarded and the pellet was resuspended in 1 ml of sterile water. Next, 100-μl aliquots were distributed in 1.5-ml sterile Eppendorf tubes and the cells were collected by centrifugation at 13,000g for 30 s. The supernatant was decanted and the cell pellet was resuspended with 326 μl of transformation mix (240 μl of PEG 3350, 36 μl of 1 M lithium acetate and 50 μl of 2 mg ml⁻¹ carrier ssDNA), plasmid DNA (500 ng of each plasmid) and sterile water to reach a final volume of 360 μl. The suspension was incubated at 42 °C for 40 min, after which it was centrifuged at 13,000g for 30 s. The supernatant was decanted, the cell pellet was resuspended in 1 ml of YPD medium and the cell suspension was incubated for 3 h at 30 °C. Cells were collected by centrifugation at 13,000g for 30 s and washed twice with 1 ml of SD medium to remove any residual YPD medium. Finally, the cell pellet was resuspended with 100 μl of SD medium, plated on solid SD medium without tryptophan and leucine and containing d-galactose and incubated at 30 °C for 3 days or until colonies were visible.

Resulting colonies were collected with a sterile 10-μl pipette tip and resuspended in 10 μl of sterile 0.02 M NaOH, boiled at 99 °C for 10 min and centrifuged for 10 s at maximum speed in a microcentrifuge. Then, 1 μl of the supernatant was used as template for PCR using the Q5 high-fidelity 2× master mix and the primer pair prCP222–prCP223 to amplify FCY1 (Supplementary Table 2). The resulting PCR product was purified using the NucleoSpin gel and PCR cleanup kit, following the manufacturer’s instructions. The final product was sequenced through Sanger sequencing. Sequence alignment was performed using the online MAFFT algorithm⁷⁷.

Base mutation assays in S.
cerevisiae

S. cerevisiae BY4741 (Δtrp1, Δleu2) cells were cotransformed with two plasmids, one bearing the specified editor variant and the other bearing either of the T sgRNAs for FCY1, ALP1 or JSN1 or an NT sgRNA (Supplementary Table 2), following identical procedures to those described above. Resulting colonies were screened through colony PCR as described above and the primer pairs prCP222–prCP223, prCP445–prCP446 and prCP441–prCP442 were used to amplify FCY1, ALP1 and JSN1, respectively (Supplementary Table 2). The resulting PCR products were sequenced through Sanger sequencing and sequence alignment was performed using the MAFFT algorithm⁷⁷.

Base mutation assays in N.
benthamiana

N. benthamiana seeds were germinated in soil and transplanted at the 1-week-old stage to 24 cell nursery flats, one plant per cell, and grown at 23 °C under a 16-h light and 8-h dark cycle in Sungro horticulture professional grow mix mixed 1:1 with Jolly gardener Pro-line C/B growing mix (Sungro).

Plasmids were used to electroporate Agrobacterium tumefaciens strain GV3101 using Bio-Rad GenePulser electroporator with the following conditions: 1.8 kV, 100 Ω and 25 µF. Single colonies were inoculated in LB medium containing spectinomycin (100 µg ml⁻¹), rifampicin (50 µg ml⁻¹) and gentamicin (50 µg ml⁻¹) for 16 h at 28 °C with orbital shaking at 200 rpm. Cultures were then centrifuged and resuspended in infiltration medium (10 mM MgCl₂ and 100 µM acetosyringone) to reach an OD₆₀₀ of ~0.1. Next, the resuspended cultures were combined in a 1:1 ratio with an A. tumefaciens strain containing p19 (a suppressor of gene silencing) and were infiltrated into the leaves of 4-week-old plants using a 1-ml needleless syringe. The infiltrated plants were then recovered overnight in the dark and grown for 7 days using the conditions mentioned above.

Next-generation sequencing in N.
benthamiana

Leaf tissues were isolated 7 days after infiltration using a standard hole punch and collected in 1.5-ml tubes containing ~100 µl of 1 mm glass beads. Disks from four leaves (one disk per leaf) were pooled to create each biological replicate. The samples were frozen at −80 °C for 24 h, after which the tissue was ground using a Vivadent shaker for 5 s followed by resuspension in CTAB buffer (1.4 M NaCl, 20 mM EDTA pH 8, 100 mM Tris-HCl pH 8 and 3% CTAB). Cellular DNA was then extracted using chloroform and isopropyl alcohol followed by a 70% ethanol wash.

The targeted region was amplified with optimized primers and PCR conditions, using an iProof high-fidelity PCR kit. The products were purified using 4 µl of ExoSAP-IT PCR product cleanup reagent (Applied Biosystems, A55242) at 37 °C for 15 min followed by inactivation at 80 °C for 15 min. A second amplification was performed with iProof polymerases to introduce unique Illumina barcodes and libraries were purified using the QIAquick gel extraction kit (Qiagen).

The concentration for each library was measured using Qubit fluorometer (Invitrogen) and equimolar amounts were pooled along with the 120 pM phiX control library corresponding to 8% of the final volume. Then, 20 μl of the pooled library was loaded into the iSeq 100 (Illumina) and the run was performed in accordance with iSeq 100 sequencing system guide. Sequencing data analysis was performed as mentioned for mammalian cells.

Mammalian cell culture and transfection

HEK293T cells were purchased from the American Type Culture Collection (CRL 11268) and U2OS^ΔTARG1 cell lines were a gift from the I. Ahel lab. Unless otherwise mentioned, all cell lines were maintained using DMEM (Life Technologies) supplemented with 10% (v/v) FBS (Corning and BANF Biotrend), 1× penicillin–streptomycin (Life Technologies) and 2 mM l-glutamine. The cultures were incubated in humidified incubators at 37 °C with 5% CO₂.

For generating the HEK293T ΔTARG1 cell line, cells were transfected with plasmids containing WT SpCas9 and TARG1 sgRNA²⁸ (Supplementary Table 2) using Lipofectamine 3000 (Invitrogen, L3000008) according to the manufacturer’s instructions. Then, 48 h after transfection, cells were diluted and seeded in 96-well plates at a density of three cells per well. Colonies were observed after 7 days and wells with single colonies were selected. Selected clones were tested for TARG1 site disruption through Sanger sequencing followed by western blotting (Supplementary Fig. 8) with anti-TARG1 antibody (Fisher Scientific, 25249-1-AP)²⁸ and anti-β-actin antibody (Life Technologies, MA5-15739-HRP) as the housekeeping control.

For templated editing assays in HEK293T (WT and ΔTARG1) cell line, 65,000 cells per well were seeded onto tissue-culture-treated 24-well plates (Corning) and incubated at 37 °C with 5% CO₂ under humidified conditions. Then, 24 h later, 50 fmol of each plasmid was cotransfected with 750 fmol of single-stranded oligodeoxynucleotide repair templates using 1.12 μl of Lipofectamine 3000 reagent and 1 μl of P3000. For base mutagenesis assays, 500 ng of each plasmid was transfected, following the same conditions as mentioned above. The medium was refreshed 24 h after transfection and cells were collected 72 h after transfection.

For base mutagenesis assays in the U2OS^ΔTARG1 cell line, 1.3 × 10⁵ cells were seeded and 1 μg of plasmid DNA, 1.5 μl of Lipofectamine 3000 reagent and 2 μl of P3000 were used for transfection. Medium change and sample collection were performed similarly to HEK293T cells.

For orthogonal R-loop assays in HEK293T^ΔTARG1 cell lines, 65,000 cells were seeded per well in 24-well plates and cotransfected after 24 h with 300 ng of SpCas9-based editor plasmids, 200 ng of SpCas9 guide plasmid, 300 ng of dSaCas9 plasmid (Addgene, 138162) and 200 ng of SaCas9 guide plasmid. Then, 1.5 µl of Lipofectamine 3000 and 2 µl of P3000 reagent were used for transfection; cell pellets were collected after 72 h.

RNA interference

For the RNA interference experiments, Dicer-substrate siRNAs were designed and purchased from Integrated DNA Technologies (TriFECTa RNAi Kit, design ID: hs.Ri.OARD1.13). All siRNA transfections were performed in HEK293T cells using Lipofectamine RNAiMAX (Invitrogen, 13778075) according to the manufacturer’s instructions. A total of 80,000 cells were seeded per well in tissue-culture-treated 24-well plates (Corning) and forward-transfected with 10 nM siRNA. After 48 h, 500 ng of each plasmid was transfected under the same conditions as described above. The medium was refreshed 24 h after plasmid transfection and cells were harvested 72 h after plasmid transfection.

The knockdown efficiency of TARG1 expression was assessed at the transcript level by real-time qPCR. Briefly, 80,000 cells were seeded and transfected with 10 nM siRNA and total RNA was extracted after 72 h using TRIzol reagent (Invitrogen) according to the manufacturer’s protocol. RNA (500 ng) was used for one-step real-time qPCR using the iTaq Universal SYBR green one-step kit (Bio-Rad, 172-5151) on a CFX96 real-time PCR detection system (Bio-Rad). The thermal cycling conditions were as follows: 50 °C for 10 min (reverse transcription), 95 °C for 1 min, followed by 40 cycles of 95 °C for 10 s and 60 °C for 30 s and a final melt curve analysis. The following primers were used for real-time qPCR: TARG1 forward, 5′-AAAGGAGACCTTTTTGCAT-3′; TARG1 reverse, 5′-GATTTAAAAGTTCTTGCACCC-3′. For each biological replicate, mRNA levels were quantified using the ΔΔC_t method, with normalization to HPRT expression and comparison to the corresponding NT siRNA control. Final values represent the mean relative expression across biological replicates.

Next-generation sequencing for mammalian cells

Genomic DNA was isolated from harvested cells using PureLink genomic DNA mini kit (Life Technologies, K182002). Specific primers were used to amplify the targeted region using Q5 high-fidelity 2× master mix through 27 cycles. The PCR product was purified using the NucleoSpin gel and PCR cleanup kit and was used as a template in KAPA HiFi HotStart ReadyMix (Roche Diagnostics, KK2602) to introduce Illumina adaptor sequences within 15 PCR cycles. The KAPA-PCR products were cleaned using Agencourt AMPure XP magnetic beads (Beckman Coulter, A63881) and 200 ng of this product was used as template for a second PCR with KAPA ReadyMix to introduce Illumina barcodes through ten PCR cycles followed by cleanup using magnetic beads as mentioned before. PCR products were screened at each step for correct fragment length using agarose gel electrophoresis. The libraries were pooled in equimolar amounts and at least 1 million reads were generated for each sample using NovaSeq 6000 and NextSeq 2000. The demultiplexed data were analyzed using CRISPResso2 (ref. ⁷⁸). Default parameters were used to perform the analysis except when quantifying indel and HDR frequencies for templated editing, in which case a plot window size of 30 was used. Allelle_frequency_table_around_sgRNA.txt files generated by CRISPResso2 were used within R scripts (https://github.com/saliba-lab/ADPr_TAE_analysis) to further quantify base mutation frequencies as the total percentage of reads containing a nucleotide different from the reference read.

Nanopore sequencing

The following steps were carried out in an amplicon-free pre-PCR area. First, 500 ng of genomic DNA was amplified using NEBNext Ultra II Q5 HiFi polymerase (New England Biolabs) with primers containing stubbers for downstream indexing. The expected amplicon length was 4.4 kb surrounding the cut site. The following PCR cycle conditions were used: denaturation at 98 °C for 30 s, followed by 25 cycles of 98 °C for 10 s, 60 °C for 30 s and 72 °C for 5 min. PCR products were purified with 0.8× solid-phase reversible immobilization beads and eluted in H₂O. Libraries were indexed and generated using the PCR barcoding expansion 1–96 (EXP-PBC096) for ligation sequencing kit (SQK-LSK114, Oxford Nanopore). Purified libraries were sequenced on a PromethION with the R10.4.1 flow cell. Read lengths were quantified using SummarizeOntDels (https://github.com/cornlab/summarizeOntDeletions)³².

Statistical analyses

For assays involving kanR reversion on solid medium (Figs. 1g and 2b), unpaired, two-tailed Welch’s t-tests were performed on log-normal data. Figure error bars display the s.d. For the nonselective editing experiment (Supplementary Fig. 3), a one-way analysis of variance was performed to test for the effect of editor–sgRNA combinations on the percentage of reads showing an SNV at the target thymidine. For the assay involving deletion strains in E. coli (Fig. 2e), unpaired, one-tailed Welch’s t-tests were performed on log-normal data. Figure error bars display the s.d. For short-read next-generation sequencing data (Figs. 3f,g and 4c–e,g), unpaired, two-tailed Welch’s t-tests were performed. Figure error bars display the s.e.m. For the editing window experiment (Fig. 4f), the median and quartiles of each group are displayed. Related P-value calculations can be found in the Source Data and Supplementary Data 1.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.