Plasmid construction
p3s-ZFD plasmids for mammalian expression were created by modifying the p3s-ABE7.10 plasmid (addgene, #113128)32 after digestion with HindIII and XhoI (NEB). The digested p3s plasmid and synthesized insert DNAs were assembled using a HiFi DNA assembly kit (NEB). All insert DNAs, which encoded MTS, ZFP (from Toolgen36, Sangamo28, and Barbas module37), split-DddA, or UGI, were synthesized by IDT.
The pTarget plasmids were designed for determining the optimal length of the spacer sequence for ZFD activity. Each pTarget plasmid, which contains two ZFP-binding sites with a spacer of variable length between them, was constructed by inserting the ZFP-binding sequences and a spacer sequence into the pRGS-CCR5-NHEJ reporter plasmid after it had been digested with two enzymes (EcoRI and BamHI, NEB), which recognize sites between the RFP and EGFP sequences.
pET-ZFD plasmids for protein production in E. coli were created by modifying the pET-Hisx6-rAPOBEC1-XTEN-nCas9-UGI-NLS plasmid (addgene, #89508)33 after digestion with NcoI and XhoI (NEB). ZFD sequences were amplified from the p3s-ZFD plasmid using PCR, and Hisx6 tag and GST tag sequences were synthesized as oligonucleotides (Macrogen). All plasmids were generated using a HiFi DNA assembly kit (NEB) to insert sequences encoding the ZFD and tag for protein purification into the digested pET plasmid.
DH5ɑ chemically competent E. coli cells were used for transformation of plasmids, and plasmids were purified with an AccuPrep Plasmid Mini Extraction Kit (Bioneer) according to the manufacturer’s protocol. The desired plasmids were selected after confirming the entire sequence with Sanger sequencing.
ZFD design scheme
In this paper, ZFDs were prepared in two ways. First, ZFDs were made by removing or attaching zinc finger protein from previously constructed ZFN. Unlike the ZFN whose optimal spacer is 5~6 bp, the optimal spacer of ZFD is ≥7 bp, so the zinc finger on the back part was removed to widen the spacer. When manufacturing zinc finger array protein, it is recommended to use four or more fingers. If there were <4 after removing one zinc finger, we added one zinc finger to the front part (Supplementary Fig. 3). Second, ZFDs were de novo assembled using a publicly available zinc finger resource (from Toolgen36). In this resource, 33 ZFP were recommended for use. In this way, since the binding ability of the ZFP array is considered critical for ZFD activity, choose the N-type or C type that can use the recommended ZFP at the DNA binding site. Finally, the generated ZFP was cloned using different original plasmids depending on the nucleus and mitochondria targets (Supplementary Figs. 4 and 7).
HEK 293 T cell culture and transfection
HEK 293 T cells (ATCC, CRL-11268) were cultured in Dulbecco’s Modified Eagle Medium (Welgene) supplemented with 10% fetal bovine serum (Welgene) and 1% antibiotic-antimycotic solution (Welgene). HEK 293 T cells (7.5 × 104) were seeded into 48-well plates. After 18–24 h, cells were transfected at 70–80% confluency with plasmids encoding left ZFD and right ZFD (500 ng each, for the full dose), or together with a pTarget plasmid (10 ng), using Lipofectamine 2000 (1.5 μL, Invitrogen). Cells were harvested at 96 h post treatment, after which they were lysed by incubation at 55 °C for 1 h, and then at 95 °C for 10 min, in 100 μL of cell lysis buffer (50 mM Tris-HCl, pH 8.0 (Sigma-Aldrich), 1 mM EDTA (Sigma-Aldrich), 0.005% sodium dodecyl sulfate (Sigma-Aldrich)) supplemented with 5 μL of Proteinase K (Qiagen).
For whole-mtDNA sequencing, HEK 293 T cells were transfected with serially diluted concentrations of plasmid or mRNA encoding ND1– or ND2-targeted mitoZFD pairs. Because more cells were required for mtDNA extraction than for analysis of the editing efficiency at defined sites, four samples, transfected under the same conditions as described above, were collected as a single sample (at four times the scale). In the manuscript, the amounts of constructs (ng) that were delivered per 7.5 × 104 cells are indicated. mtDNA was isolated from cells at 96 h post transfection.
K562 cell culture and transfection
K562 cells (ATCC, CCL-243) were maintained in RPMI 1640 supplemented with 10% fetal bovine serum (Welgene) and 1% antibiotic–antimycotic solution (Welgene).
For ZFD delivery into K562 cells by electroporation, an Amaxa 4D–Nucleofector™ X Unit system with program FF-120 (Lonza) was used. The maximum volume of substrate solution added to each sample was 2 μL when using a 16-well Nucleocuvette™ Strip. 220 pmol (for maximum capacity) or 110 pmol (for half of the maximum capacity) of each of the left and right ZFD proteins, or 500 ng of plasmid encoding left and right ZFD, were transfected into K562 cells (1 × 105). At 96 h post treatment, cells were collected by centrifugation at 100 g for 5 min, and lysed by incubation at 55 °C for 1 h, and then at 95 °C for 10 min, in 100 μL of cell lysis buffer (50 mM Tris-HCl, pH 8.0 (Sigma-Aldrich), 1 mM EDTA (Sigma-Aldrich), 0.005% sodium dodecyl sulfate (Sigma-Aldrich)) supplemented with 5 μL of Proteinase K (Qiagen).
For direct delivery of ZFD or ZFD-encoding plasmids into K562 cells, we referred to a method previously used for direct delivery of ZFN24. A mixture of left and right ZFD proteins (at a final concentration of 50 μM) or a mixture of plasmids encoding left and right ZFD (500 ng each) was diluted into serum-free medium containing 100 mM L-arginine and 90 μM ZnCl2 at pH 7.4 to a final volume of 20 μL. K562 cells (1 × 105) were centrifuged at 100 g for 5 min, and the supernatant was removed. The cells were then resuspended in the diluted ZFD solution and incubated for 1 h at 37 °C. After incubation, cells were centrifuged at 100 g for 5 min, and then resuspended in fresh culture medium. Cells were maintained at 30 °C (for a transient hypothermic condition) or 37 °C for 18 h, and then for two more days at 37 °C. Some cells were subjected to a second treatment, following the above process. Cells were analyzed 96 h after treatment.
ZFD protein expression and purification
The plasmids encoding each pair of ZFDs (Supplementary Table 7), each with a C-terminal GST tag, were transformed into Rosetta (DE3) competent cells, which were then cultured on LB-agar plates containing 50 µg/ml kanamycin. After incubation overnight, a single colony was picked and grown overnight (pre-culture) in LB broth containing 50 µg/ml kanamycin and 100 µM ZnCl2 at 37 °C. The next day, part of the pre-culture was transferred to a large volume of LB broth, which was incubated at 37 °C with shaking at 200 rpm until the absorbance, A600 nm = ~0.5–0.70. The cultures were put on ice for about 1 h, after which ZFD protein expression was induced by the addition of 0.5 mM of isopropyl β-D-1-thiogalactopyranoside (IPTG; GoldBio) and the culture was incubated at 18 °C for 14–16 h.
Protein purification steps were carried out at 0–4 °C. For cell lysis, the cells were harvested by centrifugation at 5000 g for 10 min and then resuspended in lysis buffer (50 mM Tris-HCl (Sigma-Aldrich), 500 mM NaCl (Sigma-Aldrich), 1 mM MgCl2 (Sigma-Aldrich), 10 mM 1,4-dithiothreitol (DTT; GoldBio), 1% Triton X-100 (Sigma-Aldrich), 10% glycerol, 1 mM phenylmethylsulfonyl fluoride (Sigma-Aldrich), 1 mg/ml lysozyme from chicken egg white (Sigma-Aldrich), 100 µM ZnCl2 (Sigma-Aldrich), 100 mM arginine (Sigma-Aldrich), pH 8.0). For further lysis, cells were sonicated (3 min total, 5 s on, 10 s off), after which the solution was centrifuged at 18,500 g to clear the lysate. The supernatant was then incubated with Glutathione Sepharose 4B (GE healthcare) for 1 h with gentle rotation. After this incubation, the resin-lysate mixture was loaded onto a column, which was then washed three times with wash buffer (50 mM Tris-HCl (Sigma-Aldrich), 500 mM NaCl (Sigma-Aldrich), 10 mM DTT (GoldBio), 1 mM MgCl2 (Sigma-Aldrich), 100 µM ZnCl2 (Sigma-Aldrich), 10% glycerol, 100 mM arginine (Sigma-Aldrich), pH 8.0). The bound proteins were eluted with elution buffer (50 mM Tris-HCl (Sigma-Aldrich), 500 mM NaCl (Sigma-Aldrich), 1 mM MgCl2 (Sigma-Aldrich), 40 mM glutathione (Sigma-Aldrich), 10% glycerol, 1 mM DTT (GoldBio), 100 µM ZnCl2 (Sigma-Aldrich), 100 mM arginine (Sigma-Aldrich), pH 8.0). Finally, the eluted proteins were concentrated to a concentration of ~15 ng/μL (200–240 pmol/μL, depending on the protein size) using an Amicon Ultra-4 column with a 30,000 kDa cutoff (Millipore) at 5,000 g.
In vitro deamination of PCR amplicons by ZFD
An amplicon containing the TRAC site (Supplementary Table 8) was prepared using PCR. Eight microgram of the amplicon was incubated with 2 µg of each ZFD protein (Left-G1397-N and Right-G1397-C) in NEB3.1 buffer containing 100 µM ZnCl2 for 1–2 h at 37 °C. Following the reaction, ZFD proteins were removed by incubating with 4 µL of Proteinase K solution (Qiagen) for 30 min at 55 °C, and the amplicon was purified using a PCR purification kit (MGmed). One microgram of the purified amplicon was incubated with 2 units of USER enzyme (NEB) for 1 h at 37 °C. Then, the amplicon was incubated with 4 μL of Proteinase K solution (Qiagen) and purified again using a PCR purification kit (MGmed). The product was subjected to electrophoresis on an agarose gel and imaged.
mRNA preparation
DNA templates containing a T7 RNA polymerase promoter upstream of the ZFD sequence were generated from p3s-ZFD plasmids by PCR amplification using Q5 high fidelity DNA polymerase (NEB) with forward and reverse primers (Forward: 5′-CATCAATGGGCGTGGATAG-3′, Reverse: 5′-GACACCTACTCAGACAATGC-3′). mRNAs were synthesized in vitro using an mMESSAGE mMACHINE™ T7 ULTRA Transcription Kit (Thermo Fisher). In vitro transcribed mRNAs were purified using a MEGAclear™ Transcription Clean-Up Kit (Thermo Fisher) according to the manufacturer’s protocol.
Targeted deep sequencing
Nested PCR was used to produce libraries for next-generation sequencing (NGS). The region of interest was first amplified by PCR using KAPA HiFi HotStart PCR polymerase (Roche). To generate NGS libraries, amplicons were amplified again using TruSeq DNA-RNA CD index-containing primers to label each fragment with adapter and index sequences. Final PCR products were purified using a PCR purification kit (MGmed) and sequenced using a MiniSeq sequencer (Illumina) with a GenerateFASTQ workflow. Primer sequences for targeted deep sequencing are listed in Supplementary Table 10. Substitution and indel frequencies from targeted deep sequencing data were calculated with source code (https://github.com/ibs-cge/maund, written by BotBot Inc.).
Whole mitochondrial genome sequencing
For whole mitochondrial genome sequencing, three steps were required: mtDNA extraction from isolated mitochondria, NGS library generation, and NGS. First, 3 × 105 HEK 293 T cells were trypsinized and collected by centrifugation (500 g, 4 min, 4 °C) 96 h after transfection with ND1– or ND2-targeted mitoZFD pairs. Then, cells were washed with ice-cold phosphate-buffered saline (Welgene), and collected again by centrifugation. The supernatant was removed, and the mitochondria were isolated from cultured cells using the reagent-based method of the Mitochondria Isolation Kit for Cultured Cells (Thermo Fisher) according to the manufacturer’s protocol. mtDNA was extracted from isolated mitochondria with a DNeasy Blood & Tissue Kit (Qiagen). To generate an NGS library from the extracted mtDNA, we used an Illumina DNA Prep kit with Nextera™ DNA CD Indexes (Illumina). Finally, the libraries were pooled and loaded onto a MiniSeq sequencer (Illumina). The average sequencing depth was >50.
Analysis of mitochondrial genome-wide DNA editing
To analyze NGS data from whole mitochondrial genome sequencing, we referred to a method previously used for DNA off-target analysis of TALE-DdCBE3. First, we aligned the Fastq files to the GRCh38.p13 (release v102) reference genome using BWA (v.0.7.17), and generated BAM files with SAMtools (v.1.9) by fixing read pairing information and flags. Then, we used the REDItoolDenovo.py script from REDItools (v.1.2.1)38 to identify, among all cytosine and guanines in the mitochondrial genome, the positions with conversion rates ≥1%. We excluded positions with conversion rates ≥50% in all samples, regarding these as single-nucleotide variations in the cell lines. We also excluded the on-target sites for each ZFD treatment. We considered the remaining positions to be off-target sites and counted the number of edited C/G nucleotides with an editing frequency ≥1%. We averaged the conversion rates at each base position in the off-target sites to calculate the average C/G to T/A editing frequency for all C/Gs in the mitochondrial genome. Specificity ratios were calculated by dividing the average on-target editing frequency by the average off-target editing frequency. Mitochondrial genome-wide graphs were created by plotting the conversion rates at on-target and off-target sites.
Data visualization
GraphPad Prism 8, Adobe Illustrator CS6, Microsoft Excel 2016, and PowerPoint 2016 were used for generating figures and tables.
Reporting summary
Further information on research design is available in the Nature Research Reporting Summary linked to this article.
