A multi-kingdom genetic barcoding system for precise clone isolation – Nature Biotechnology

Plasmids

Oligonucleotides were chemically synthesized by FASMAC, Integrated DNA Technologies or Eurofins Genomics. All oligonucleotides and cloning procedures used to construct the plasmids in this study are listed in Supplementary Table 2. We used QUEEN (v.1.2.0) (https://github.com/yachielab/QUEEN) to design each plasmid construction and generate annotated plasmid files in QUEEN’s GenBank (gbk) file format, embedding the full construction procedure (see Supplementary Table 2). A QUEEN gbk file acts as a quine code that enables retrieving the plasmid construction process that generates the same plasmid map in the gbk format⁷¹. We believe that providing these QUEEN gbk files fulfils the requirement for reporting reproducible plasmid construction protocols. We also provided natural language descriptions for all the plasmid construction protocols in the QUEEN gbk files. Users can retrieve the protocols by executing ‘QUEEN –protocol_description –input [gbk file]’ in a QUEEN-installed environment. A custom QUEEN wrapper that generated all QUEEN-generated gbk files is also available at https://github.com/yachielab/CloneSelect_v1/tree/main/QUEEN. Accordingly, we do not include plasmid construction protocols in this paper. All plasmid DNA sequences were confirmed by Sanger sequencing. The representative plasmids are available from Addgene along with their QUEEN gbk files, as agreed upon with Addgene.

Common methods

Lentivirus preparation

Packaging

HEK293T cells were plated either in a 10 cm cell culture dish at a density of ~2 × 10⁶ cells in 10 ml of culture medium or in six-well cell culture plate wells at a density of ~2 × 10⁵ cells per well in 2 ml of culture medium, 1 day before plasmid transfection.

For virus packaging in a 10 cm dish, 3.0 µg of the transgene vector, 2.25 µg of psPAX2 (Addgene, no. 12260), 0.75 µg of pMD2.G (Addgene, no. 12259) and 18 µl of 1 mg ml⁻¹ PEI MAX (Polysciences, no. 24765-100) were dissolved in 1,000 µl of 1× PBS and added to the cell culture. For packaging in a six-well plate, 489 ng of the transgene plasmid, 366.7 ng of psPAX2, 122.3 ng of pMD2.G and 2.93 µl of 1 mg ml⁻¹ PEI MAX were dissolved in 300 µl of 1× PBS and added to the culture.

The culture medium was replaced with fresh medium 1 day after transfection. Transfected cells were then incubated for an additional 48–72 h. The recombinant lentivirus supernatant was collected and filtered through 0.22 µm sterile syringe filters. The lentivirus samples were aliquoted in 500–1,000 µl volumes into 1.5 ml tubes and stored at −80 °C.

Virus concentration

To increase the viral infection titre, collected virus samples were concentrated using a polyethylene glycol (PEG)-based method⁷² with PEG 6000 (Wako, no. 169-09125) or with Lenti-X Concentrator (Takara, no. 631231).

For concentration with PEG 6000, approximately 10 ml of the recombinant virus sample was combined with 2.55 ml of 50% w/v PEG 6000, 1.085 ml of 4 M NaCl and 1.365 ml of 1× PBS in a 50 ml tube. The mixture was rotated continuously at 4 °C for 90 min, then centrifuged at 4,000g and 4 °C for 20 min. The supernatant was discarded, and the virus pellet was resuspended in 1.1 ml of Opti-MEM (Gibco, no. 31985062) by pipetting and vortexing until fully dissolved, achieving a tenfold concentration of the virus sample.

Virus concentration using Lenti-X Concentrator followed the manufacturer’s protocol, with the virus dissolved in Opti-MEM (Gibco, no. 31985062) for a tenfold or 15-fold concentration. The concentrated virus samples were stored at −80 °C.

Preparing microscope imaging samples

All live-cell imaging was conducted using a BZ-X710 (Keyence), InCellAnalyzer 6000 (GE Healthcare) or IX83 (Olympus) with a ×4, ×10 or ×20 objective lens. The contrast and brightness of images obtained in a single experimental batch were uniformly adjusted using ImageMagick (v.7.1.0-20) or Fiji (v.1.0).

HEK293T cells and mouse ES cells were analyzed with Hoechst staining. For HEK293T cells, 25 µl of 0.1 mg ml⁻¹ Hoechst 33342 (Invitrogen, no. H3570) dissolved in DMEM was directly added to each well of 24-well cell culture plates 3 days after transfection for nuclear counterstaining. The specimens were incubated at room temperature (18–25 °C) for 10 min, after which the culture medium was removed. Cells were gently washed with 500 µl of fresh DMEM and filled with 500 µl of fresh DMEM before imaging. For mouse ES cells, 5.0 µg ml⁻¹ Hoechst 33342 dissolved in cell culture medium was directly added to each well and incubated at room temperature for 10 min before imaging.

Flow cytometry analysis

Cells were detached with 0.25% w/v trypsin-EDTA (Wako, no. 201-18841), incubated at 37 °C for 5 min, collected into a 1.5 ml tube or a 96-well round-bottom plate and centrifuged at 100g at room temperature for 5 min. After aspirating the supernatant, cell pellets were gently resuspended in 150–500 µl of ice-cold FACS buffer (2% FBS in 1× PBS). Samples were immediately placed on ice until flow cytometry analysis.

Flow cytometry analysis was performed using a BD FACSVerse cell analyzer (BD Biosciences) or CytoFLEX flow cytometer (Beckman Coulter). Samples were gently mixed by pipetting or vortexing immediately before analysis, and approximately 10,000–20,000 raw events were acquired per sample. Data analysis was conducted with custom R scripts using flowWorkspace (v.0.5.40) (https://github.com/RGLab/flowWorkspace), flowCore (v.1.11.20) (https://github.com/RGLab/flowCore) and CytoExploreR (v.1.1.00) (https://github.com/DillonHammill/CytoExploreR) or with the Python package FlowCytometryTools (v.0.5.0) (https://github.com/eyurtsev/FlowCytometryTools). The codes are available at https://github.com/yachielab/CloneSelect_v1/tree/main/FACS.

High-throughput sequencing

All amplicon sequencing libraries were combined with a 20–30% PhiX spike-in DNA control (Illumina, no. FC-110-3001) to enhance cluster generation on the flow cell. Libraries were sequenced using Illumina MiSeq (MiSeq v.3 150-cycle kit no. MS-102-3001 or 300-cycle kit no. MS-102-3003) or HiSeq 2500 (TruSeq rapid SBS kit v.2 no. FC-402-4022). Base calling was performed with bcl2fastq2 (v.2.20.0) to generate FASTQ files. Detailed sequencing conditions for each library and NCBI Sequence Read Archive IDs for each raw FASTQ file are provided in Supplementary Table 3.

Barcode identification and analysis

In barcode identification of each different barcoding system, sequencing reads were aligned to the constant sequences of the library structure using NCBI BLAST+ (v.2.6.0)⁷³ with the blastn-short option to identify sample indices for demultiplexing and barcode sequences. For the clone isolation experiments, a barcode allowlist was generated by identifying barcode sequences present in both the plasmid DNA library and the genomic DNA library. Sequencing errors were corrected using Starcode (v.1.4) (https://github.com/gui11aume/starcode) with a maximum Levenshtein distance threshold of four, merging minor barcodes into major ones.

Barcode counts in each sample were normalized by the total barcode count. Barcode frequencies for each cell or DNA pool sample were estimated by averaging frequencies across replicates, where applicable. The barcode sequence and frequency data generated in this study are provided in Supplementary Table 1.

Statistical analysis

Statistical tests were conducted using R (v.4.2.0 and v.4.3.1). Specific details for each test are provided in the corresponding figure legends. Additionally, the statistical methods and associated P values used in this study are listed in Supplementary Table 4.

Experiments using HEK293T cells

Cell culture

HEK293Ta and HEK293T Lenti-X cells were purchased from GeneCopoeia (no. LT008) and Takara (no. 632180), respectively. Cells were cultured in DMEM (Sigma-Aldrich, no. 11965084) supplemented with 10% FBS (Gibco, no. 16000044) and 1% penicillin–streptomycin (Wako, no. 168-23191) at 37 °C with 5% CO₂ in a cell culture incubator. Cells were detached and passaged using 0.25 w/v% trypsin-EDTA (Wako, no. 203-20251) once they reached 70–90% confluency. For microscopic imaging of HEK293T cells with Hoechst 33342 (Invitrogen, no. H3570) counterstain, 100–200 µl of Collagen-I (Nippi, no. PSC-1-100-100) diluted in 5 mM acetic acid was added to each cell culture plate well and incubated for 30 min at 37 °C. The collagen-coated plate wells were washed with 100–200 µl of 1× PBS before use. Cells were regularly tested for mycoplasma contamination.

Barcode plasmid pool preparation

CloneSelect C→T barcode library

To generate the CloneSelect C→T barcode library, a semi-random oligonucleotide pool, SI#679, encoding 5′-CCGWSNSWSNSWSNSWSNSNGTG-3′, was first chemically synthesized (Supplementary Table 2). This sequence includes the antisense strand of the 5′-CGG-3′ PAM sequence, followed by a quadruple repeat of WSNS (where W = A or T; S = G or C) and a mutated start codon (GTG). The WSNS repeat prevents the formation of additional start and stop codons upstream of the reporter. An EGFP coding sequence was then amplified from pLV-eGFP (Addgene, no. 36083) in 25 separate 50 µl PCR reactions, each containing 1 ng µl⁻¹ of pLV-eGFP template plasmid, 1.25 µl of 20 µM SI#679 oligonucleotide pool as the forward primer, 1.25 µl of 20 µM SI#680 as the common reverse primer, 0.5 µl of Phusion High-fidelity DNA Polymerase (NEB, no. M0530), 10 µl of 5× Phusion HF Buffer (NEB, no. B0518S) and 5 µl of 2.5 mM deoxynucleotide triphosphates (dNTPs; Takara, no.4025). The thermal cycling conditions were as follows: 98 °C for 30 s; 30 cycles of 98 °C for 10 s, 72 °C for 10 s and 72 °C for 60 s; with a final extension at 72 °C for 5 min.

The amplified fragment was digested with DpnI (NEB, no. R0176) for 1 h at 37 °C, pooled into a single 1.5 ml tube and purified using the FastGene PCR/Gel Extraction Kit (Nippon Genetics, no. FG-91302). The purified fragment was then subjected to overnight digestion with EcoRI-HF (NEB, no. R3101S) and XbaI (NEB, no. R0145S) at 37 °C, followed by another purification with the FastGene PCR/Gel Extraction Kit. To obtain a highly complex lentiviral plasmid pool, we performed five ligation reactions using PCR strip tubes, each containing ~30 fmol of EcoRI-XbaI-digested pLVSIN-CMV-Pur backbone plasmid (Takara, no. 6183), ~300 fmol of the insert fragment, 2.5 µl of T4 DNA Ligase (NEB, no. M0202) and 5 µl of 10× T4 DNA Ligase Reaction Buffer (NEB, no. B0202) in a total volume of 50 µl. Reaction samples were incubated at room temperature for 2 h and then purified using the FastGene PCR/Gel Extraction Kit.

The ligation samples were used to transform NEB Stable Competent E. coli cells (NEB, no. C3040I) in five separate reactions, each with 1,250 ng of the ligation sample in 200 µl of competent cells, following the manufacturer’s high-efficiency transformation protocol. After a 1 h outgrowth in SOC medium (NEB, no. B9020) at 37 °C, cells were spun down and plated across 25 LB agar plates containing 100 µg ml⁻¹ ampicillin (Wako, no. 014-23302). Colonies that formed on each plate after overnight incubation at 37 °C were scraped with 1–2 ml of double-distilled water (ddH₂O). The cells collected from the plates were pooled into a flask and incubated in 200–300 ml of LB liquid medium with 100 µg ml⁻¹ ampicillin (Wako, no. 014-23302) overnight at 37 °C. The transformation sample was plated with a 500-fold dilution in triplicate, and the library’s complexity was estimated to be ~6.8 × 10⁵. The plasmid library was then purified using the NucleoBond Midi-prep Kit (Macherey-Nagel, no. 740410) and stored at −20 °C.

We isolated 16 random clones and verified the presence of the expected barcode inserts through triple restriction enzyme digestion with BsrGI-HF (NEB, no. R3575S), ClaI (NEB, no. R0197S) and PvuI-HF (NEB, no. R3150S), confirming that 16 out of 16 clones contained the desired inserts. To generate the mini-pool library for proof-of-concept assays in HEK293T cells, we sequenced barcode regions from 96 isolated clones by Sanger sequencing with primer SI#471. After excluding three clones with mixed sequencing spectra in the barcode region, the remaining barcoded plasmids were pooled in equimolar ratios and subjected to high-throughput sequencing and lentiviral packaging.

CloneSelect C→T Pool-10000 barcode library

To generate the CloneSelect C→T Pool-10000 barcode library, 100 ng of the original 700K library plasmid pool was re-transformed into 10 µl of NEB Stable Competent E. coli cells (NEB, no. C3040I). This transformation was controlled to confer approximately 10,000 colonies. The collected colonies were pooled and cultured overnight in 5 ml LB liquid medium containing 100 µg ml⁻¹ ampicillin (Wako, no. 014-23302) at 30 °C. Plasmid DNA was extracted using the GeneJET Plasmid Miniprep Kit (Thermo Fisher Scientific, no. K0502) and stored at −20 °C until use.

CaTCH and ClonMapper Pool-10000 libraries

The CaTCH and ClonMapper Pool-10000 libraries were constructed using Golden Gate Assembly⁷⁴ with the same protocol. To prepare an insert fragment pool, two single-stranded DNA oligonucleotide pools containing a random 19-mer nucleotide sequence were synthesized by Integrated DNA Technologies and annealed to generate sticky-end overhangs (underlined): 5′-CACCCNNNNNNNNNNNNNNNNNNNG-3′ and 5′-AAACCNNNNNNNNNNNNNNNNNNNG-3′ for CaTCH; 5′-CACCGNNNNNNNNNNNNNNNNNNG-3′ and 5′-AAACCNNNNNNNNNNNNNNNNNNC-3′ for ClonMapper (Supplementary Table 2). Equal volumes of top and bottom strand oligonucleotide pools were combined for phosphorylation and annealing in a 30 µl reaction volume in an eight-strip PCR tube. The reaction mixture included 3 µl of 10× T4 PNK Buffer (Takara, no. 2021A), 1.5 µl of T4 PNK (Takara, no. 2021A) and 3 µl each of 100 µM top and bottom strand oligonucleotide pools. The mixture was incubated with the following thermal cycling conditions: 37 °C for 30 min, 95 °C for 5 min, 70 cycles of 12 s at 95 °C with a ramp down of 1 °C per cycle, and then maintained at 25 °C. The annealed oligonucleotide pool was diluted to 1/10 with ddH₂O and used for Golden Gate Assembly with the appropriate lentiviral cloning backbone (pLV-CS-307 and lentiTRACE-hU6-Puro for CaTCH and ClonMapper, respectively). The Golden Gate Assembly reaction mix was prepared in a 12.5 µl volume in an eight-strip PCR tube, consisting of 1 µl of insert, 1.25 µl of 10× T4 DNA Ligase Reaction Buffer (NEB, no. B0202S), 0.625 µl of 2 mg ml⁻¹ BSA (NEB, no. B9000S), 0.5 µl of T4 DNA Ligase (Nippon Gene, no. 317-00406), 0.5 µl of BsmBI (NEB, no. R0580), 1.25 µl of 25 mM ATP (NEB, no. P0756S) and 12.5 ng of the backbone plasmid. The assembly reaction underwent the following thermal cycling conditions: 15 cycles of 37 °C for 5 min and 20 °C for 5 min, followed by 55 °C for 30 min, then held at 4 °C.

Following assembly, 3 µl of the product was transformed into NEB Stable Competent E. coli cells (NEB, no. C3040I) using the high-efficiency transformation protocol. After 1 h of outgrowth in SOC medium (NEB, no. B9020) at 30 °C, cells were spun down and plated on LB agar plates containing 100 µg ml⁻¹ ampicillin (Gibco, no. 11593027). This transformation was controlled to confer approximately 10,000 colonies. After overnight incubation at 30 °C, colonies on each plate were scraped into 1–2 ml of LB medium containing 100 µg ml⁻¹ ampicillin and pooled in a 5 ml tube. The collected cell samples were further incubated overnight with 4–6 culture tubes, each with 5 ml of LB liquid medium with 100 µg ml⁻¹ ampicillin at 30 °C. Plasmid DNA was extracted using the GeneJET Plasmid Miniprep Kit (Thermo Fisher Scientific, no. K0502) and stored at −20 °C.

To confirm library quality, a random subset of clones was isolated and subjected to genotyping PCR with primer pairs SI#157–SI#766 for the ClonMapper library or SI#2040–SI#330 for the CaTCH library. Barcode sequences of the clones were further verified by Sanger sequencing.

Barcode sequencing library preparation

CloneSelect C→T mini-pool library

To identify barcodes in the CloneSelect C→T mini-pool library by high-throughput sequencing, ~10 ng of plasmid DNA (approximately 1.0 × 10⁹ molecules) was used as a PCR template. For identifying barcodes in the initial barcoded HEK293Ta cell population, genomic DNA was purified using NucleoSpin Tissue (Macherey-Nagel, no. 740952) according to the manufacturer’s protocol, and 119 ng of extracted genomic DNA (4 × 10⁴ molecules, 400-fold of the estimated barcode complexity) was used as a PCR template.

The sequencing libraries were prepared using a two-step PCR method. The first-round PCR was performed in a 20 µl reaction containing template DNA, 0.5 µl each of 20 µM forward (SI#682) and reverse (SI#683) primers, 0.2 µl of Phusion High-Fidelity DNA Polymerase (NEB, no. M0530), 4 µl of Phusion HF Buffer (NEB, no. B0518S), 2 µl of 2 mM dNTPs (Takara, no. 4025) and 0.6 µl of 100% DMSO (NEB, no. 12611P). The thermal cycling conditions were as follows: 98 °C for 10 s; 30 cycles of 98 °C for 10 s, 61 °C for 10 s and 72 °C for 30 s; followed by a final extension at 72 °C for 5 min. Each PCR product was size-selected using 2% agarose gel and purified with the PCR/Gel Extraction Kit (Nippon Genetics, no. FG-91302).

To add Illumina sequencing adaptors and custom indices, the second-round PCR was performed on each first-round PCR product in a 20 µl reaction containing 2.5 ng of the first PCR product, 1 µl each of 10 µM P5 and P7 custom index primers, 0.2 µl of Phusion High-Fidelity DNA Polymerase (NEB, no. M0530), 4 µl of Phusion HF Buffer (NEB no. B0518S), 2 µl of 2 mM dNTPs (Takara, no. 4025) and 0.6 µl of 100% DMSO (NEB, no. 12611P). The thermal cycling conditions were as follows: 98 °C for 10 s; 20 cycles of 98 °C for 10 s, 61 °C for 10 s and 72 °C for 30 s; followed by a final extension at 72 °C for 5 min. Custom indices for the second-round PCR products are listed in Supplementary Table 3. Each second-round PCR product was size-selected and purified using the PCR/Gel Extraction Kit (Nippon Genetics, no. FG-91302). Sequencing samples were pooled, quantified by qPCR using the Kapa Library Quantification Kit Illumina (Kapa Biosystems, no. KK4824) and analyzed by paired-end sequencing using Illumina MiSeq.

CloneSelect C→T, CaTCH and ClonMapper Pool-10000 libraries

To identify barcodes in each Pool-10000 plasmid library by high-throughput sequencing, ~1 pg of plasmid DNA was used as a PCR template. To identify barcodes in each barcoded HEK293Ta cell population, genomic DNA was purified using the NucleoSpin Tissue Kit (Macherey-Nagel, no. 740952) according to the manufacturer’s protocol, and a total of ~2 µg of genomic DNA was used as a PCR template. Sequencing libraries were prepared using a two-step PCR method.

The first-round PCR reaction mixture was split into 20 subreactions and distributed into 20 wells of a 96-well plate for each of the two replicates. Each 25 µl subreaction contained template DNA, 1.0 µl each of 10 µM forward and reverse primers, 0.5 µl of Phusion High-Fidelity DNA Polymerase (NEB, no. M0530), 5 µl of 5× Phusion HF Buffer (NEB, no. B0518S) and 0.5 µl of 10 mM dNTPs (NEB, no. N0447S). For CloneSelect C→T, the primer pair and the thermal cycle conditions were the same as described above. For CaTCH, the primer pair was CS-310-PS1.0-FW4 and CS-310-PS2.0-RV1, and the thermal cycling conditions were as follows: 98 °C for 10 s; 10 cycles of 98 °C for 10 s, 67 °C for 15 s and 72 °C for 30 s; followed by 20 cycles of 98 °C for 10 s and 72 °C for 1 min; with a final extension at 72 °C for 7 min. For ClonMapper, the primer pair was PS1.0-hU6-FW5 and Scaffold-PS2.0-RV5, and the thermal cycling conditions were as follows: 98 °C for 10 s; 30 cycles of 98 °C for 10 s, 67 °C for 15 s and 72 °C for 30 s; with a final extension at 72 °C for 7 min. PCR products were pooled and purified with a 1.8× volume of Agencourt AMPure XP magnetic beads (Beckman Coulter, no. A63881) following the manufacturer’s protocol.

To add Illumina sequencing adaptors and custom indices, the second-round PCR was performed on each first-round PCR product in a 25 µl reaction containing 10 ng of the first PCR product, 0.75 µl each of 10 µM P5 and P7 custom index primers, 0.5 µl of Kapa HiFi DNA Polymerase (Kapa Biosystems, no. KK2101), 5 µl of 5× Kapa HiFi Fidelity Buffer (Kapa Biosystems, no. KK2101) and 0.75 µl of 10 mM dNTPs (NEB, no. N0447S). The thermal cycling conditions were as follows: 95 °C for 5 min; 10 cycles of 98 °C for 20 s, 67 °C for 15 s and 72 °C for 1 min; followed by 10 cycles of 98 °C for 20 s and 72 °C for 1 min; with a final extension at 72 °C for 1 min. Custom indices for the second-round PCR products are listed in Supplementary Table 3. The second-round PCR products were purified using a 1.8× volume of Agencourt AMPure XP magnetic beads (Beckman Coulter, no. A63881). Sequencing samples were pooled, quantified by qPCR using the Kapa Library Quantification Kit Illumina (Kapa Biosystems, no. KK4824) and analyzed by paired-end sequencing using Illumina MiSeq.

Sorted cells

For amplicon sequencing-based barcode identification for low-volume cells obtained by barcode-specific clone isolation in the CloneSelect C→T mini-pool assays, a cell lysate was prepared for each sample as a PCR template. Cells in eight-strip PCR tubes were first incubated with 2.0 µl of lysis buffer containing 600 mM KOH, 10 mM EDTA and 100 mM dithiothreitol. The samples were then neutralized with 2.0 µl of neutralization buffer composed of 0.4 µl of 1 M Tris-HCl and 1.6 µl of 3 M HCl. For the first-round PCR, 2.0 µl of the cell lysate was used as the template. Although no visible bands were observed on gel electrophoresis for the first-round PCR products, the PCR product of the expected size was isolated using 2% agarose gel, purified and eluted in 15 µl of ddH₂O with the PCR/Gel Extraction Kit (Nippon Genetics, no. FG-91302).

The PCR products were quantified using the Quant-iT PicoGreen dsDNA Assay Kit (Thermo Fisher Scientific, no. P7589) and the Infinite 200 PRO plate reader (TECAN) with Tecan i-control software (v.1.10.4.0). For the second-round PCR, 2.0 ng of the first-round PCR product was used as the template. Custom indices assigned to the second-round PCR products are provided in Supplementary Table 3.

The second-round PCR products were size-selected and purified with the PCR/Gel Extraction Kit (Nippon Genetics, no. FG-91302). The samples were pooled into a DNA LoBind 1.5 ml tube (Eppendorf, no. 13-698-791), quantified by qPCR using the Kapa Library Quantification Kit Illumina (Kapa Biosystems, no. KK4824) and analyzed by paired-end sequencing on an Illumina MiSeq.

For the Pool-10000 assays, following cell sorting, genomic DNA was extracted using the NucleoSpin Tissue Kit (Macherey-Nagel, no. 740952). Sequencing libraries were constructed using the protocols described for the barcoded Pool-10000 cell populations.

Cell barcoding

Cells were seeded in six-well cell culture plates at a density of ~2 × 10⁵ cells per well in 2 ml of culture medium for barcoding of cells with single barcodes, a 10 cm dish at a density of ~2 × 10⁶ cells per dish in 10 ml of culture medium for establishing the CloneSelect C→T mini-pool cell population and a 15 cm dish at a density of ~1 × 10⁷ cells per dish for establishing the Pool-10000 pool cell populations. The next day, a total of 500–1,000 µl transduction mix containing 2 µg ml⁻¹ Polybrene (Sigma-Aldrich, no. TR-1003), recombinant lentivirus and cell culture medium was applied to each well alongside non-virus controls. The following day, the culture medium was replaced with fresh medium containing 2.0 µg ml⁻¹ puromycin (Gibco, no. A1113803) or 5.0 µg ml⁻¹ blasticidin S (Wako, no. 029-18701) to select infected cells over 2–5 days. Barcoded cell populations were maintained with 500–1,000 coverages while expanding and passaging.

After drug selection, cell viability was measured using CellTiter-Glo (Promega, no. G7570) according to the manufacturer’s protocol, and luminescence was quantified with the Infinite 200 PRO plate reader (TECAN). Background luminescence from wells without cells was subtracted from all readings.

For each condition, the infection rate was calculated as the fraction of surviving cells compared to non-selective controls. Samples with an infection rate close to but not exceeding 0.1 were used in subsequent analyses, in which most selected cells were expected to contain a single viral integration based on Poisson statistics.

Preparing Pool-10000 cell pools

After barcoding cells with the CloneSelect C→T and CaTCH Pool-10000 libraries, background EGFP reporter expression was observed in some cells. To eliminate possible false positives before the experiment, EGFP⁻ cells were first collected by flow cytometry cell sorting while maintaining the original barcode complexity. Approximately ~8 × 10⁶ cells, representing an average of 800 clones per barcode, were sorted for both libraries using the MoFlo Astrios (Beckman Coulter). Following sorting and expansion to ~90% confluency, genomic DNA was purified for the barcode sequencing analysis using the NucleoSpin Tissue Kit (Macherey-Nagel, no. 740952) according to the manufacturer’s protocol.

Selecting target clones for isolation from the Pool-10000 cell populations

For each CloneSelect C→T, CaTCH and ClonMapper Pool-1000 cell population, 16 clones of a diverse range in abundance were arbitrarily selected for isolation. In each population, the barcode abundance rates were grouped into four bins: (0.01, 0.02], (0.025, 0.05], (0.05, 0.1] and (0.1, 1.0]. From each bin, four target clones were randomly chosen. If a bin contained fewer than four clones, additional clones were randomly selected from the next higher bin to reach a total of 16 target barcodes for testing. The targeted clones in each assay are listed in Supplementary Table 1.

Reporter activation

For all experiments delivering a reporter activation reagent to a cell sample of a single barcode, cells were seeded in 24-well cell culture plates at a density of ~5 × 10⁴ cells per well in 500 µl of culture 1 day before transfection. A total of 400 ng of plasmids with a 3:1 mass ratio of a Cas9 effector or decoy plasmid to gRNA plasmid, 1.2 µl of 1 mg ml⁻¹ PEI MAX (Polysciences, no. 24765) and 100 µl of 1× PBS were combined, incubated for 5–10 min at room temperature and applied to each well (for CaTCH, 300 ng of a decoy plasmid PLVSIN-CMV-Pur and 100 ng of a gRNA plasmid were used). For the dose-dependent reporter activation assay with different Target-AID expression plasmids, transfections were performed in 24-well plates with plasmid amounts per well ranging from 50 to 800 ng. The PEI MAX volume was adjusted to 3 µl per 1 µg of plasmid.

For isolating a target clone from the CloneSelect C→T mini-pool, cells were seeded in six-well plates at a density of ~2 × 10⁵ cells per well in 2,000 µl of culture medium 1 day before transfection. A total of 800 ng of a plasmid encoding both Target-AID and a gRNA were combined with 2.5 µl of 1 mg ml⁻¹ PEI MAX (Polysciences, no. 24765) and 200 µl of 1× PBS, then applied to each well after a 5–10 min incubation at room temperature.

For isolating a target clone from each Pool-10000 cell population, cells were cultured in 15 cm dishes with a seeding density of approximately 2–4 × 10⁶ cells. Then, 1 day before transfection, CloneSelect C→T and CaTCH Pool-10000 cells were seeded in 10 cm dishes at a density of ~2 × 10⁶ cells per dish in 10 ml of culture medium. ClonMapper Pool-10000 cells were seeded in six-well plates at a density of ~2 × 10⁵ cells per well in 2 ml of culture medium.

The following day, CloneSelect C→T Pool-10000 cells were co-transfected with 5,250 ng of the Target-AID expression plasmid (pRS0035) and 1,750 ng of the barcode-targeting gRNA plasmid using 22.5 µl of 1 mg ml⁻¹ PEI MAX (Polysciences, no. 24765) and 300 µl of 1× PBS. CaTCH Pool-10000 cells were co-transfected with 5,250 ng of a decoy plasmid (pcDNA3.1 V5-HisA) and 1,750 ng of the barcode-targeting gRNA plasmid using 22.5 µl of 1 mg ml⁻¹ PEI MAX and 300 µl of 1× PBS. ClonMapper Pool-10000 cells were co-transfected with 550 ng of the dCas9-VPR expression plasmid (pLV-CS-282 v2) and 450 ng of the barcode-targeting reporter plasmid using 3 µl of 1 mg ml⁻¹ PEI MAX and 100 µl of 1× PBS. The transfection mix was incubated for ~5 min at room temperature and then applied to each sample.

Flow cytometry cell sorting

In the isolation of a target clone from the CloneSelect C→T mini-pool, cells were detached using 0.25% w/v trypsin-EDTA (Wako, no. 201-18841) 4 days after transfection of the reporter activation reagents, incubated at 37 °C for 5 min, collected into a 1.5 ml tube and centrifuged at 100g at room temperature for 5 min. Cells were then resuspended in a 5 ml polystyrene round-bottom tube (FALCON) containing 150–500 µl of 1% FBS in 1× PBS and immediately placed on ice until sorting. Sorting was conducted using the BD FACSJazz (BD Biosciences) in 1.0 drop single sort mode. Cells were initially gated using FSC-A and SSC-A, with the gate for EGFP⁺ cells defined by selecting those with high FITC-A intensities, which were absent in a control sample transfected with Target-AID and non-target gRNA plasmids. EGFP⁺ cells were sorted into eight-strip PCR tubes (Nippon Genetics, no. FG-018WF), each containing 2.5 µl of 1× PBS. For optimal recovery, the collection tube’s cell destination position was manually adjusted for each sample. Sorted cells were immediately placed on an ice-cold 96-well aluminum block. Although the rate of EGFP⁺ cells varied across samples, approximately 50–600 EGFP⁺ cells were recovered per experiment.

In the Pool-10000 experiments, cell samples of different activation reagents were each detached using 1× PBS, detached with 0.25% trypsin-EDTA, phenol red (Gibco no. 25200072) 3 days after transfection and combined into 50 ml tubes for each replicate group. The pooled cell samples were resuspended in FACS buffer (2% FBS in 1× PBS) and kept on ice before sorting.

Cell sorting was performed on a MoFlo Astrios (Beckman Coulter). Owing to a low frequency (~0.01%) of EGFP⁺ cells for CloneSelect C→T and CaTCH, an initial enrichment sort was performed for ~1.4 × 10⁸ cells to increase EGFP⁺ cells to 20–30%. The EGFP-enriched cells were then sorted again using the purity sort mode to obtain ~5 × 10³ EGFP⁺ cells per sample. For ClonMapper, cells were sorted directly using a purity sort mode to obtain ~3 × 10⁵ EGFP⁺ cells per sample. The EGFP⁺ gate was defined using a non-transfected cell sample for each sample.

The raw data for cell sorting is available at https://github.com/yachielab/CloneSelect_v1/tree/main/FACS/Raw_flow_data.

Experiments using mouse ES cells

Cell culture

Under approval from the Institutional Animal Care Committee of the University of Tokyo (RAC180003), mouse ES cells were derived from embryos of a 129(+Ter)/SvJcl (female mouse) × C57BL/6NJcl (male mouse) cross and maintained in DMEM low glucose (Sigma-Aldrich, no. D6046-500ML) supplemented with 1% penicillin–streptomycin (Gibco, no. 15140122), 1% MEM non-essential amino acids (Wako, no. 139-15651), 1% GlutaMAX supplement (Gibco, no. 35050061), 1% sodium pyruvate (Gibco, no. 11360070), 15% FBS (Gibco, no. 16000044), 100 µM 2-mercaptoethanol (Wako, no. 131-14572), 1,000 units per ml ESGRO Recombinant Mouse LIF Protein (Millipore, no. ESG1107), 3.0 µM CHIR99021 (GSK-3 inhibitor) (Wako, no. 038-23101) and 1.0 µM PD0325901 (MEK inhibitor) (Tocris, no. 4423). Before seeding cells, 0.1% gelatin (Sigma-Aldrich, no. G9391) in 1× PBS (Takara, no. T9181 or Gibco, no. 70011044) was added to each well, covering the entire surface, and then aspirated after 1 h at 37 °C. Cells were cultured at 37 °C with 5% CO₂ in a cell culture incubator, and the cell culture medium was replaced at least every 2 days. Cells were regularly tested for mycoplasma contamination.

Cells with stably integrated Target-AID

The mouse ES cell line with stably integrated Target-AID was established by electroporation using the NEPA21 Super Electroporator (Nepa Gene). After detaching cells from culture plate wells, ~2 × 10⁶ cells were mixed with 100 µl of Opti-MEM (Gibco no. 31985062), 2.0 µg of pNM1325 and 0.7 µg of a Super piggyBac transposase vector (SBI, no. PB210PA-1), then transferred to an electroporation cuvette (Nepa Gene, no. EC-002S). The electroporation was performed with two poring pulses of positive polarity at 115 V for 5 ms, with 50 ms intervals and a 10% decay rate. Five transfer pulses were then applied for both positive and negative polarities at 20 V for 50 ms, with 50 ms intervals and a 40% decay rate. After electroporation, cells were transferred to a 10 cm culture dish with fresh medium, which was replaced with fresh medium again 1 day post electroporation. Then, 2 days after electroporation, the medium was replaced with medium containing 5 µg ml⁻¹ of blasticidin S (Wako, no. 029-18701) to select cells with stable integration. Cells were incubated for about 2 weeks in the selection medium.

Transfection

Cells were seeded in 48-well cell culture plates at a density of ~6 × 10⁴ cells in 200 µl of culture medium. For each transfection reaction, 200 ng of a gRNA plasmid was diluted in 20 µl of Opti-MEM (Gibco, no. 31985062). Separately, 0.6 µl of Lipofectamine 2000 (Invitrogen no. 11668019) was combined with 19.4 µl of Opti-MEM to form the transfection mix. The plasmid solution and transfection mix were then combined and applied to each well after a 5 min incubation at room temperature.

Barcode plasmid pool preparation

The scCloneSelect barcode library was prepared similarly to the CloneSelect C→T barcode library. An EGFP coding sequence was first amplified from pLV-CS-112 (Addgene, no. 131127) by PCR using the semi-random oligonucleotide pool SI#679 as the forward primer and RS#244 as the reverse primer. The PCR was performed in 25 separate 40 µl reactions, each containing 0.12 µl of 10 ng µl⁻¹ pLV-CS-112 template plasmid, 2 µl each of forward and reverse primers, 0.6 µl of Phusion High-fidelity DNA Polymerase (NEB, no. M0530), 8 µl of 5× Phusion HF Buffer (NEB, no. B0518S) and 3.2 µl of 2.5 mM dNTPs (NEB, no. N0447). The thermal cycling conditions were as follows: 98 °C for 30 s; followed by 30 cycles of 98 °C for 10 s, 65 °C for 10 s and 72 °C for 60 s; with a final extension at 72 °C for 5 min.

The amplified barcode-EGFP fragment was pooled into a single 1.5 ml tube, digested with 12.5 µl of DpnI (NEB, no. R0176) at 37 °C for 1 h and size-selected using the FastGene PCR/Gel Extraction Kit (Nippon Genetics, no. FG-91302). The purified product was then subjected to overnight digestion with EcoRI-HF (NEB, no. R3101S) and XbaI (NEB, no. R0145) at 37 °C and purified again using the FastGene PCR/Gel Extraction Kit. For backbone preparation, 25 µg of the pRS193 lentiviral cloning backbone plasmid was digested with EcoRI-HF (NEB, no. R3101S) and XbaI (NEB, no. R0145) overnight at 37 °C and then size-selected with the FastGene PCR/Gel Extraction Kit.

The ligation reaction was prepared by mixing 1.25 µg of the digested backbone, 320 ng of the purified insert, 25 µl of T4 DNA Ligase (Nippon Gene, no. 317-00406) and 25 µl of 10× T4 DNA Ligase Reaction Buffer (NEB, no. B0202) in a total volume of 250 µl, followed by overnight incubation at 16 °C. The ligation mixture was then transformed into NEB Stable Competent E. coli cells (NEB, no. C3040I) across 17 reactions, each containing 4 µl of the ligation sample and 50 µl of competent cells, following the manufacturer’s high-efficiency transformation protocol. After 1 h of outgrowth in SOC medium (NEB, no. B9020) at 37 °C, cells were centrifuged and plated across 15 LB agar plates containing 100 µg ml⁻¹ ampicillin (Wako, no. 014-23302). Colonies that formed on each plate after overnight incubation at 37 °C were scraped with 1–2 ml ddH₂O. The collected cell samples were pooled and further incubated in 200–300 ml of LB liquid medium with 100 µg ml⁻¹ ampicillin (Wako, no. 014-23302) overnight at 37 °C. A 300-fold diluted transformation sample was plated in duplicate on agar, estimating the barcode complexity at ~1.5 × 10⁵. The plasmid library was purified using the NucleoBond Midi-prep Kit (Macherey-Nagel, no. 740410) and stored at −20 °C.

We isolated 20 random clones and verified fragment insertion by genotyping PCR with primer pair RS#147 and SI#514, confirming the expected insertion in 17 out of 20 clones. From these, we selected six clones (including three without expected genotyping bands) for double digestion with EcoRI-HF (NEB, no. R3101S) and BamHI-HF (NEB, no. R3136S), followed by Sanger sequencing using primers SI#514 and RS#147 for the uptag and dntag, respectively. All tested clones contained the expected uptag and dntag inserts.

Barcode sequencing library preparation

Uptag–dntag combination reference database

To establish the uptag–dntag combination reference database for the barcoded mouse ES cell population, genomic DNA was first extracted from ~1 × 10⁵ cells using NucleoSpin Tissue (Macherey-Nagel no. 740952) following the manufacturer’s protocol. Sequencing libraries were prepared using a two-step PCR method, with 50 ng of genomic DNA per PCR reaction.

The first-round PCR was performed in a 20 µl reaction containing template DNA, 0.7 µl each of 10 µM forward (SI#682) and reverse (RS#250) primers, 0.2 µl of Phusion High-Fidelity DNA Polymerase (NEB, no. M0530), 4.5 µl of Phusion HF Buffer (NEB, no. B0518S) and 1.6 µl of 2.5 mM dNTPs (NEB, no. N0447). The thermal cycling conditions were as follows: 98 °C for 10 s; 15 cycles of 98 °C for 10 s, 60 °C for 10 s and 72 °C for 2 min; followed by a final extension at 72 °C for 5 min. Each PCR product was size-selected using 2% agarose gel, purified and eluted in 20 µl of ddH₂O with the PCR/Gel Extraction Kit (Nippon Genetics, no. FG-91302).

To add Illumina sequencing adaptors and custom indices, the second-round PCR was performed in a 20 µl reaction using a 20-fold dilution of the first-round PCR product, 0.7 µl each of 10 µM P5 and P7 custom index primers, 0.2 µl of Phusion High-Fidelity DNA Polymerase (NEB, no. M0530), 4.5 µl of Phusion HF Buffer (NEB, no. B0518S) and 1.6 µl of 2.5 mM dNTPs (NEB, no. N0447). The thermal cycling conditions were as follows: 98 °C for 10 s; 20 cycles of 98 °C for 10 s, 60 °C for 10 s and 72 °C for 30 s; followed by a final extension at 72 °C for 5 min. Custom indices for the second-round PCR products are listed in Supplementary Table 3. The second-round PCR products were size-selected and purified using the PCR/Gel Extraction Kit (Nippon Genetics, no. FG-91302). Sequencing samples were pooled, quantified by qPCR with the Kapa Library Quantification Kit Illumina (Kapa Biosystems, no. KK4824) and analyzed by paired-end sequencing using Illumina MiSeq.

Sorted cells

For cells sorted after gRNA-dependent barcode-specific clone isolation, a cell lysate was prepared for each sample as a PCR template. The sequencing library of each sample was generated by modifying the two-step PCR method for the uptag–dntag combination reference database.

Cell samples were first expanded in 96-well culture plate wells until confluent. After aspirating the culture medium, 20 µl of 50 mM NaOH was added to each well, and the contents were transferred to a 96-well PCR plate for direct cell lysis. The samples were then heated at 95 °C for 15 min and cooled on ice, followed by neutralization with 2.0 µl of 1 M Tris-HCl (pH 8.0).

The first-round PCR was performed in a 40 µl reaction, with 3.5 µl of cell lysate as the template. The second-round PCR was performed in a 20 µl reaction, using a tenfold dilution of the first-round PCR product as the template. Custom indices assigned to the second-round PCR products are provided in Supplementary Table 3. The second-round PCR products were size-selected and purified using the GeneJET Gel Extraction Kit (Thermo Fisher Scientific, no. K0691). Sequencing samples were pooled into a DNA LoBind 1.5 ml tube (Eppendorf, no. 0030108051), quantified by qPCR with the Kapa Library Quantification Kit Illumina (Kapa Biosystems, no. KK4824) and analyzed by paired-end sequencing using Illumina HiSeq 2500.

Reamplification of dntags from Drop-seq library

To increase the sensitivity of identifying dntags associated with single-cell transcriptome profiles, the DNA region encoding dntags and cell IDs were selectively reamplified from the intermediate Tn5 transposon-fragmented sample of the Drop-seq process and sequenced separately.

The reamplification PCR was performed in a 20 µl reaction containing 1 ng of template DNA (quantified using TapeStation with High Sensitivity D5000 ScreenTape; Agilent, nos. 5067-559 and 5067-5593), 0.7 µl each of 20 µM forward primer P5-TSO_Hybrid⁴³ and reverse primer SI#682, 0.2 µl of Phusion High-Fidelity DNA Polymerase (NEB, no. M0530), 4.5 µl of 5× Phusion HF Buffer (NEB, no. B0518) and 1.6 µl of 2.5 mM dNTPs (NEB no. N0447). The thermal cycling conditions were as follows: 95 °C for 30 s; 30 cycles of 98 °C for 30 s, 60 °C for 10 s and 72 °C for 2 min; followed by a final extension at 72 °C for 5 min. The first-round PCR product was purified and eluted in 20 µl of ddH₂O using the GeneJET Gel Extraction Kit (Thermo Fisher Scientific, no. K0691).

The second-round PCR was performed in a 20 µl reaction using a tenfold dilution of the first-round PCR product, 0.7 µl each of 10 µM P5 and P7 custom dual index primers, 0.2 µl of Phusion High-Fidelity DNA Polymerase (NEB, no. M0530), 4.5 µl of 5× Phusion HF Buffer (NEB, no. B0518) and 1.6 µl of 2.5 mM dNTPs (NEB, no. N0447). The thermal cycling conditions were as follows: 95 °C for 30 s; 15 cycles of 98 °C for 10 s, 65 °C for 10 s and 72 °C for 2 min; followed by a final extension at 72 °C for 5 min. Custom indices for the second-round PCR products are listed in Supplementary Table 3. The second-round PCR products were size-selected using 2% agarose gel, purified with the GeneJET Gel Extraction Kit (Thermo Fisher Scientific, no. K0691), pooled, quantified by qPCR using the Kapa Library Quantification Kit Illumina (Kapa Biosystems, no. KK4824) and analyzed by paired-end sequencing using Illumina HiSeq 2500.

Cell barcoding

To introduce a single barcode, cells with or without stably integrated Target-AID were seeded in six-well cell culture plates at a density of ~2 × 10⁵ cells per well in 2 ml of culture medium 1 day before transduction. A recombinant virus sample with a 10–100 µl volume was thawed on ice, mixed with 1.5 µl of 8 µg ml⁻¹ Polybrene (Sigma-Aldrich, no. TR-1003) and 1.5 ml of fresh culture medium and then applied to the cells. To select transduced cells, the culture medium was replaced with a fresh medium containing 1.0 µg ml⁻¹ puromycin (Gibco no. A1113803) 2 days after infection, followed by an additional 3 days of incubation. Surviving cells were detached, and cell counts were measured using an automated cell counter (BioRad TC20). The infection rate was calculated as the fraction of surviving cells compared to the non-selective control condition. Samples with an infection rate close to but not exceeding 0.1 were used in subsequent analyses.

For the barcoding of a cell population, cells with stably integrated Target-AID were seeded in six-well cell culture plate wells at a density of ~2 × 10⁵ cells per well with 2 ml of culture medium 1 day before transduction. The following day, cells were transduced with 500 µl of a 15-fold concentrated barcoding lentivirus pool using the same transduction protocol and selected 2 days after infection. For the downstream proof-of-principle differentiation and clone isolation assays, a clonal population bottleneck was created by seeding ~1,000 cells in a single six-well plate and culturing them for 10 days.

Mouse ES cell differentiation assay

The barcoded cell population with the clonal complexity bottleneck was then divided as follows: ~1 × 10⁴ cells were seeded into culture medium with LIF and 2i (1.0 µM PD0325901; Tocris, no. 4423 and 3.0 µM CHIR99021; Wako, no. 038-23101) (LIF+2i+), ~1 × 10⁴ cells were seeded into culture medium without LIF or 2i (LIF−2i−), two samples of ~1 × 10⁵ cells each were set aside to establish the uptag–dntag combination reference database and five replicates of ~1 × 10⁵ cells were stored at −80 °C in CELLBANKER 1 freeze medium (ZENOAQ, no. 11910). Then, 4 days later, cells in both the LIF+2i+ and LIF−2i− conditions were subjected to scRNA-seq.

Drop-seq

scRNA-seq was performed by Drop-seq with devices manufactured by Dolomite Bio according to the manufacturer’s protocol. Microfluidic devices were fabricated by YODAKA. Cell samples were prepared at a concentration of ~2 × 10⁵ cells per ml for analysis.

Sequencing libraries were prepared following the original Drop-seq protocol⁴³. In brief, after emulsion breakage and reverse transcription, ‘single-cell transcriptomes attached to microparticles’ (STAMPs) were washed and treated with Exonuclease I (NEB, no. M0293L). Approximately 2,000 STAMPs were used for the whole cDNA amplification of each sample. Following second-strand synthesis, library DNA was purified with AMPure XP beads (Beckman Coulter, no. A63881), quantified using a TapeStation with High Sensitivity D5000 ScreenTape (Agilent, nos. 5067-5592 and 5067-5593) and fragmented with Tn5 transposon using the Nextera XT DNA Library Preparation Kit (Illumina, no. FC-131-1024) as per the manufacturer’s protocol. The fragmented sequencing library was purified with AMPure XP beads (Beckman Coulter, no. A63881) and quantified again using the TapeStation with High Sensitivity D5000 ScreenTape (Agilent, nos. 5067-5592 and 5067-5593). Each library’s average size was confirmed to be ~500 bp. Multiple scRNA-seq libraries were pooled and subjected to high-throughput sequencing using Illumina MiSeq or HiSeq 2500. The sequencing library index information is provided in Supplementary Table 3.

RT–PCR

The transcription of polyadenylated scCloneSelect barcode products was assessed by PCR with reverse transcription (RT–PCR) and gel electrophoresis. Total RNA was extracted using the ISOSPIN Cell & Tissue RNA Kit (Nippon Gene, no. 314-08211) according to the manufacturer’s instructions. The RNA was then treated with DNase I (Takara, no. 2270B) to eliminate residual DNA and purified again using the ISOSPIN Cell & Tissue RNA Kit (Nippon Gene, no. 314-08211).

First-strand cDNA was synthesized using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, no. 4368814) in a 10 µl reaction volume containing 5 µl of DNase I-treated RNA (~1 µg), 0.5 µl of 100 µM oligonucleotide dT primer SI#4, 0.5 µl of MultiScribe Reverse Transcriptase, 1 µl of 10× RT buffer, 0.4 µl of 100 mM dNTPs and 0.5 µl of RNase Inhibitor (Applied Biosystems, no. N8080119). The thermal cycling conditions were as follows: 25 °C for 10 min, 37 °C for 12 min and 85 °C for 5 min.

The transcription of the target barcode was then analyzed by PCR alongside a GAPDH control. Each PCR reaction was conducted in a 20 µl volume, containing 2 µl of 50-fold diluted first-strand cDNA, 2.8 µl total of either the primer pair SI#116–SI#7 to amplify the dntag or the primer pair RS#507–RS#508 to amplify GAPDH, 0.2 µl of Phusion DNA Polymerase (NEB, no. M0530S), 4 µl of 5× Phusion HF Buffer (NEB, no. B0518) and 1.6 µl of 2.5 mM dNTPs (NEB, no. N0447). The thermal cycling conditions were as follows: 98 °C for 30 s; 30 cycles of 98 °C for 10 s, 60 °C for 10 s and 72 °C for 30 s; followed by a final extension at 72 °C for 5 min. The PCR products were analyzed on a 2% agarose gel.

Flow cytometry cell sorting

Each cell sample was expanded in a 10 cm cell culture dish 3 days after transduction with a query gRNA. Cells were detached with 0.25% w/v trypsin-EDTA (Gibco, no. 25200072), incubated at 37 °C for 5 min, collected into a 1.5 ml tube and centrifuged at 100g at room temperature for 5 min. The cells were then resuspended to approximately 1 × 10⁶ cells in PBS containing 2% FBS and transferred to a 5 ml polystyrene round-bottom tube (Falcon, no. 352054). The cell suspension was immediately placed on ice until sorting.

Sorting was conducted using MoFlo Astrios EQ Cell Sorter (Beckman Coulter). Cells were initially gated using FSC-A and SSC-A, with the gate for EGFP⁺ cells set to include those with high FITC-A intensities, which were absent in a non-transduced control sample. EGFP⁺ cells were single-cell sorted into 96-well plate wells, while the remaining cells were sorted in bulk into a single well of a 96-well plate, each containing 100 µl of mouse ES cell culture medium. Approximately 100–1,000 EGFP⁺ cells were recovered per experiment, except for clone 153, for which no EGFP⁺ cells above the gating threshold were observed.

The raw data for cell sorting is available at https://github.com/yachielab/CloneSelect_v1/tree/main/FACS/Raw_flow_data.

Barcode analysis

To identify uptag and dntag barcodes in a cell population, sequencing reads were first demultiplexed, and cutadapt (v.4.1) (https://github.com/marcelm/cutadapt) was used to extract uptag and dntag sequences located between their 20 bp upstream and downstream constant sequences. Extracted uptags and dntags were filtered with a Q-score threshold of 30, then clustered and further filtered by length (17 bp for uptags and 30 bp for dntags) using bartender-1.1 (https://github.com/LaoZZZZZ/bartender-1.1)⁷⁵.

In constructing the uptag–dntag combination reference database, redundant uptag–dntag pairs with either uptag or dntag found in more abundant pairs were discarded. For mapping dntags to the uptag–dntag database, symspellpy (v.6.7) (https://github.com/mammothb/symspellpy) was used to find the match with the shortest edit distance. If multiple dntags with the same edit distance were found, the dntag with the highest frequency in the database was selected.

To analyze uptag frequencies in cell populations following gRNA-dependent barcode-specific EGFP reporter activation and flow cytometry sorting, sequencing reads were mapped to the uptag–dntag database, and uptag read counts were obtained using bartender-1.1 and symspellpy (v.6.7).

The codes used for the barcode identification are available at https://github.com/yachielab/CloneSelect_v1/tree/main/Barcode_identification/scCloneSelect.

Drop-seq data analysis

After sample demultiplexing of Illumina sequencing reads, FASTQ files were processed with Drop-seq Tools (v.2.5.1) (https://github.com/broadinstitute/Drop-seq) for base quality filtering, adaptor trimming and extraction of cell ID and unique molecular identifier sequences.

Picard (v.2.18.14) (https://github.com/broadinstitute/picard) was used to convert BAM files back to FASTQ files for subsequent steps. Filtered reads were aligned using STAR (v.2.7) (https://github.com/alexdobin/STAR)⁷⁶ with the mm10 reference genome.

Differential gene expression and clustering analyses were conducted using Seurat (v.3) (https://github.com/satijalab/seurat)⁷⁷. Cells were filtered based on thresholds of Feature_RNA > 200, nFeature_RNA < 2500 and percent.mt < 5, and gene expression profiles were normalized using the Seurat::sctransform function before clustering.

To identify dntags for analysis, we performed an initial Drop-seq run and determined dntags based on the cumulative read count distribution of cell IDs, with a threshold set at the knee point using the Python package kneed (v.0.8.1) (https://github.com/arvkevi/kneed). For higher sensitivity in mapping dntag distributions to single-cell transcriptome data, we also sequenced the reamplified dntag library and used the dntag–uptag combination reference database to identify cell ID and dntag associations, as described for the scCloneSelect library preparation. When multiple dntags were associated with a single cell ID, the dntag with the highest unique molecular identifier count was selected.

The codes are available at https://github.com/yachielab/CloneSelect_v1/tree/main/Drop-seq.

Experiments using human PS cells

Cell culture

The CA1 human PS cell line was used with approval from the Canadian Institutes of Health Research Stem Cell Oversight Committee. CA1 human PS cells were cultured in mTeSR Plus medium (STEMCELL Technologies, no. 100-0276) in a humidified incubator at 37 °C with 21% O₂ and 5% CO₂. Culture plates were coated with Geltrex LDEV-Free Reduced Growth Factor Basement Membrane Matrix (Gibco, no. A1413201). To prepare the Geltrex working solution, DMEM/Nutrient Mixture F-12 (DMEM/F-12) (Gibco, no. 11320033) was diluted 1:100 with Geltrex. A sufficient volume of this solution was added to each well, covering the surface, and was aspirated after 1 h of incubation at 37 °C before plating cells. The cell culture medium was replaced every other day after cell seeding. Cells were routinely passaged as medium-sized clumps. After aspirating the medium, ReLeSR (STEMCELL Technologies, no. 05872) was added, and cells were incubated at room temperature for about 1 min before a second aspiration. Cells were then placed in the incubator for 4–5 min, fresh medium was added and cells were dissociated by gentle pipetting. The cells were then plated and returned to the incubator.

For single-cell passaging, TrypLE Express (Gibco, no. 12604021) was used. The cells were incubated for 4 min before adding fresh medium to stop the action of TrypLE Express. Cells were collected in centrifuge tubes, dissociated by pipetting and filtered through a 40 µm cell strainer (Sarstedt, no. 83.3945.040) to remove clumps. Tubes were centrifuged at 300–400g for 5 min, and the supernatant was aspirated. Pellets were resuspended in fresh medium supplemented with 10 µM ROCK inhibitor Y-27632 (Tocris Bioscience, no. 1254) for 24 h to support single-cell survival.

For culturing H1 human PS cells, we used StemFit AK02N medium (REPROCELL AHS, no. RCAK02N), with Y-27632 (Cayman, no. 10005583) added for 1–2 days after plating. Culture plates were coated with recombinant Laminin-511 E8 fragment using iMatrix-511 Silk (MAX, no. 892021).

The Center for iPS Cell Research and Application (CiRA) Ethics Committee, an internal committee at Kyoto University’s CiRA, approved our research plan for human ES cell research (CiRA21-03) and recombinant DNA experiments (240283). The WiCell line H1 (WA01) was used under agreements 10-WO-0098, 23-W0713 and 24-W0434.

Cells were regularly tested for mycoplasma contamination.

Cells with stably integrated Target-AID

To establish a human PS cell line with stably integrated Target-AID, CA1 cells were seeded in 24-well cell culture plates at a density of ~5 × 10⁴ cells per well in 1 ml of culture medium 1 day before transfection. The transfection mix was prepared by combining 450 ng of pNM1325 (CAGp-Target-AID-2A-Blast), 50 ng of a hyperactive piggyBac transposase plasmid, 1 µl of Lipofectamine Stem Transfection Reagent (Invitrogen, no. STEM00001) and 49 µl of Opti-MEM (Gibco, no. 31985062) and was applied to the wells after 10 min of incubation. The following day, the culture medium was replaced with fresh medium to remove residual transfection reagent. Then, 3 days post transfection, the medium was replaced with fresh medium containing 5 µg ml⁻¹ of blasticidin S to initiate selection for 24 h. An additional two-day selection was performed until cells reached confluency, at which point they were passaged into a new culture plate. A final selection round was conducted to ensure the selection of the cells.

Cell barcoding

For the introduction of a single barcode, cells with or without stably integrated Target-AID, cells were seeded in six-well cell culture plates at a density of ~1 × 10⁵ cells per well in 2 ml of culture medium 1 day before transduction. For transduction, recombinant virus samples with a volume of 10–100 µl were thawed on ice, mixed with 1.5 µl of 8 µg ml⁻¹ Polybrene (Sigma-Aldrich, no. TR-1003) and 1.5 ml of fresh culture medium and then applied to the cells. After 48 h of infection, the culture medium was replaced with fresh medium containing 1.0 µg ml⁻¹ puromycin (Gibco, no. A1113803) for 3 days. The reporter-integrated cells were then dissociated into single cells and subjected to flow cytometry sorting to enrich EGFP⁻ cells. The sorted cells were maintained in StemFit AK02N culture medium (REPROCELL, no. RCAK02N).

For the barcoding of the H1 cell population, cells were seeded at a density of ~2.1 × 10⁴ cells per cm² 1 day before transduction. The following day, freshly prepared medium containing 50 µl of the barcoded virus library and 2 µl of 8 mg ml⁻¹ Polybrene (Nacalai Tesque, no. 12996-81) was added to each well. After 48 h, the culture medium was replaced with fresh medium containing 1.0 µg ml puromycin (Life Technologies, no. A1113802) for 3 days to select for transduced cells. Following puromycin selection, EGFP⁻ cells were enriched by flow cytometry cell sorting using the BD FACS Aria (BD Biosciences).

Naive induction

Barcoded cells were seeded at a density of ~1.6 × 10⁴ cells per cm² with iMatrix-511 silk (MATRIXOME, no. 387-10131) in StemFit AK02N (Ajinomoto, no. RCAK02N). After 48 h, naïve induction was initiated with cRM-1 + Y culture medium (designated as day 0), consisting of NDiff 227 (Takara Bio, no. Y40002) supplemented with 1 µM PD0325901 (Tocris, no. 4192), 10 ng ml⁻¹ Recombinant Human LIF (Peprotech, no. 300-05), 1 mM valproic acid (Sigma-Aldrich, no. P4543) and 10 µM Y-27632 (Cayman, no. CAY-10005583-50). Then, 2 days later, the culture medium was switched to PXGL + Y medium, composed of NDiff 227 (Takara Bio, no. Y40002) with added 1 µM PD0325901 (Tocris, no. 4192), 10 ng ml⁻¹ Recombinant Human LIF (Peprotech, no. 300-05), 2 µM Go 6983 (Tocris, no. 2285), 2 µM XAV-939 (Selleck, no. S1180) and 10 µM Y-27632 (Cayman, no. CAY-10005583-50). Cells were passaged using TrypLE Express Enzyme (Invitrogen, no. 12604021) and Enzyme Free Cell Dissociation Solution (Sigma-Aldrich, no. S-014-B) and cultured in the same medium for 23–25 days.

Reporter activation

Reporter activation assays using CA1 human PS cells

To activate the reporter of a barcoded CA1 human PS cell sample with stably integrated Target-AID, we used the Neon Transfection System (Invitrogen, no. MPK5000) to deliver the gRNA plasmid by electroporation. Cells were detached from culture plate wells, and ~1 × 10⁵ cells were mixed with 100 µl of Neon Resuspension Buffer (Invitrogen, no. MPK10096) and 2.0 µg of gRNA plasmid. Electroporation was performed with the following settings: 1,200 V, 30 ms, single pulse.

Reporter activation assays using H1 human PS cells

To activate the reporter of a barcoded H1 human PS cell sample, Target-AID and gRNA expression plasmids were co-delivered by electroporation. Cells were detached from culture plate wells, and ~1 × 10⁵ cells were mixed with 100 µl of Neon Resuspension Buffer (Invitrogen, no. MPK10096), 3.0 µg of Target-AID plasmid and 3.0 µg of gRNA plasmid. Electroporation was performed with the following settings: 1,200 V, 20 ms, two pulses.

Elite clone isolation from the barcoded H1 human PS cell clone population

To isolate a target clone from the barcoded H1 human PS cell population, Target-AID and gRNA expression plasmids were also co-delivered by electroporation using the Neon Transfection System (Invitrogen, no. MPK5000). Cells were detached with Accutase (Sigma-Aldrich, no. A6964-500ML), and ~2.0 × 10⁶ cells were transferred to a 1.5 ml tube. The cells were washed once with 1× D-PBS (-) (Nacalai Tesque, no. 14249-24) and resuspended in 100 µl of Neon Resuspension Buffer (Invitrogen, no. MPK10096) containing 9 µg of the Target-AID plasmid and 6 µg of the gRNA plasmid. Electroporation was performed with the following settings: 1,200 V, 20 ms, two pulses.

Flow cytometry cell sorting

Cell samples were washed with 1× D-PBS (-) and detached using 2 ml of Accutase (Sigma-Aldrich, no. A6964-500ML) to create a single-cell suspension. Cells were resuspended in FACS buffer composed of 450 ml MilliQ water, 50 ml 10× Hanks’ Balanced Salt Solution (no calcium, no magnesium, no phenol red) (Invitrogen, no. 14185052) and 5 g BSA (Sigma-Aldrich, no. A2153-100G) and kept on ice for 30 min.

Immunostaining was conducted on ice with antibodies in FACS buffer for 30 min. Flow cytometry and cell sorting were performed using the BD LSR Fortessa or FACS Aria II systems (BD Bioscience). The following antibodies were used: anti-human SUSD2 antibody (PE) (Biolegend, no. 327406; 1:200 dilution), CD24 monoclonal antibody (APC) (Thermo Fisher Scientific, no. 17-0247-42; 1:200 dilution), TROP2 antibody, anti-human, REAfinity (Biotin) (Miltenyi Biotec, no. 130-115-054; 1:200 dilution), mouse anti-human CD249 (BV421) (BD Bioscience, no. 744872; 1:200 dilution) and APC streptavidin (Biolegend, no. 405207; 1:1,000 dilution). Data analysis was conducted with FlowJo (v.10.7.2).

The raw data for cell sorting is available at https://github.com/yachielab/CloneSelect_v1/tree/main/FACS/Raw_flow_data.

Trophoblast differentiation

The protocol for trophoblast differentiation was previously established and described⁷⁸. In brief, H1 naïve stem cells were seeded at a density of ~2.0 × 10⁴ cells per cm² onto iMatrix-511 silk in NDiff 227 medium supplemented with 2 µM A 83-01 (Tocris, no. 2939), 2 µM PD0325901 and 10 ng ml⁻¹ BMP-4 (R&D, no. 314-BP-500). The following day, the medium was replaced with NDiff 227 supplemented with 2 µM A 83-01, 2 µM PD0325901 and 1 µM JAK Inhibitor I (Calbiochem, no. 420099). On day 3, cells were detached using Accutase (Sigma-Aldrich, no. A6964-500ML), immunostained with anti-human TROP2 (Miltenyi Biotec, no. 130-115-054) and anti-human CD249 (BD Bioscience, no. 744872) and then sorted using the BD LSR Fortessa or FACS Aria II systems (BD Bioscience). Trophoblast marker genes used in this study were curated from a previous report⁷⁸.

qPCR

HAVCR1⁺/ENPEP⁺ cells were subjected to total RNA extraction using the Quick-RNA Kit Micro-Prep (ZYMO, no. R1051). Total RNA (0.5 µg) was reverse-transcribed into cDNA with an oligonucleotide dT primer using SuperScript IV (Invitrogen, no. 18090050). qPCR was conducted using PowerUP SYBR Green Master Mix (Applied Biosystems no. A25743), following the manufacturer’s instructions. Results were analyzed with QuantStudio Design & Analysis Software (Thermo Fisher Scientific, v.1.4.1). Cycle threshold values were normalized to GAPDH to calculate the relative expression of trophoblast marker genes. Primer pairs used for qPCR are listed in Supplementary Table 2.

RNA-seq

Sequencing library preparation

RNA-seq libraries were prepared from 1 ng of total RNA using the SMART-Seq HT Kit (Takara, no. Z4436N) following the manufacturer’s instructions. Sequencing libraries were pooled with PhiX Control (v.3) (Illumina, no. FC-110-3001) and sequenced using Illumina NovaSeq 6000 with paired-end sequencing.

Data processing

RNA-seq reads were trimmed to remove adaptor sequences and low-quality bases using cutadapt (v.4.1) (https://github.com/marcelm/cutadapt). The trimmed reads were then aligned to the human reference genome (hg38) with STAR (v.2.7.10a)⁷⁶. Read counts for each gene were obtained from the resulting BAM files using HTSeq (v.2.0.2).

Differential gene expression analysis was performed with DESeq2 (v.1.34.0)⁷⁹ in R (v.4.1.1). The differentially expressed genes were identified from the DESeq2 output with an adjusted P value threshold of 0.05. To obtain normalized gene expression data for z-score standardization, a regularized log transformation was applied using the rlog function.

For visualizing read mapping in Integrated Genomics Viewer (IGV; v.2.16.2)⁸⁰, BAM files were converted to BigWig format using the bamCoverage command in deepTools (v.3.5.4)⁸¹. Gene expression matrices were further processed for hierarchical clustering and visualization with the pheatmap package (v.1.0.12) (https://github.com/raivokolde/pheatmap) in R (v.4.3.1).

GSEA

To identify robust gene expression signatures in the isolated clones, clone 006, clone 034, clone 116, clone 216 and clone 332 were grouped as case samples, while the wild-type and barcoded wild-type samples were grouped as control samples. GSEA was performed on the log-transformed gene expression data using GSEApy (v.1.1.1)⁸². GSEA was conducted against the ‘GO_Biological_Process_2023’ gene set using the gseapy.gse function, and the enriched Gene Ontology terms were filtered with a false discovery rate threshold of 0.1. The Gene Ontology term database was obtained from the Enrichr website⁸³.

The resulting GSEA data was converted to a graph structure using the Gene Ontology database go-basic.obo (release date 2024-01-17), obonet (v.1.0.0) (https://github.com/dhimmel/obonet) and networkx (v.3.2.1) (https://github.com/networkx/networkx) on Python (v.3.10.0). Cytoscape (v.3.10.1)⁸⁴ was used for visualization.

EM-seq

Sequencing library preparation

EM-seq libraries were constructed according to the original protocol⁵⁴. Genomic DNA was purified from ~1.0 × 10⁶ input cells using the Wizard Genomic DNA Purification Kit (Promega, no. A1120). DNA concentration was quantified with the Qubit 1× dsDNA High Sensitivity Assay Kit (Thermo Fisher Scientific, no. Q33231), and 200 ng of genomic DNA was mixed with 20 pg of unmethylated lambda DNA and 1 pg of CpG-methylated pUC19 as internal controls.

The mixed DNA was fragmented using a Covaris E220 focused-ultrasonicator with the following settings: peak incident power at 175 W, duty factor at 10%, cycles per burst at 140, treatment time of 90 s and temperature range between 0 and 40 °C. DNA fragment size was verified with the High Sensitivity D5000 ScreenTape Assay (Agilent, no. 5067-5588) on the 4200 TapeStation System (Agilent), confirming predominant fragment sizes between 150 and 600 bp.

Library preparation followed the standard NEBNext Enzymatic Methyl-seq Kit protocol (no. E7120). After end-repair, A-tailing and EM-seq adaptor ligation, 5-methylcytosines and 5-hydroxymethylcytosines were oxidized with TET2 and deaminated with APOBEC. The library was then PCR-amplified and purified. Quantification was conducted using the High Sensitivity D5000 ScreenTape Assay Kit (Agilent, no. 5067-5588) on the 4200 TapeStation System (Agilent) and Qubit 1× dsDNA High Sensitivity Assay Kit (Thermo Fisher Scientific, no. Q33231). All EM-seq libraries were pooled with PhiX Control (v.3) (Illumina, no. FC-110-3001) and sequenced using Illumina NovaSeq 6000.

Data processing

EM-seq adaptor sequences were trimmed, and low-quality reads were discarded using Trim Galore (v.0.6.10) (https://github.com/FelixKrueger/TrimGalore). The processed reads were aligned to the human reference genome (hg38) with Bismark (v.0.24.1)⁸⁵. The aligned reads were deduplicated using the deduplicate_bismark command, and methylated bases were called with the bismark_methylation_extractor command, applying the options –ignore 2, –ignore_r2 2 and –ignore_3prime_r2 3 to minimize methylation biases near the read ends.

BedGraph files for methylated bases were generated using the bismark2bedGraph command with the options –CX and –cutoff 3. Methylation reports for each nucleotide context were computed using the coverage2cytosine command with the –CX option. To visualize the methylation profile in IGV, we extracted cytosines in the CpG context and calculated the proportion of methylated cytosines in 500 bp bins with read counts of >20. The data were then converted into BigWig format using bedGraphToBigWig (v.2.10) (https://github.com/ENCODE-DCC/kentUtils). Unless otherwise specified, default settings were applied for all commands.

Methylation profiling was conducted with methylKit (v.0.9.7)⁸⁶. Cytosines in the CpG context with a minimum coverage of three reads were extracted, and the reference genome was divided into 1,000-bp windows. Bins with fewer than ten reads were discarded. The binned CpG profiles were subjected to differential methylation analysis between the sorted clones and their corresponding parental samples using the calculateDiffMeth function in methylKit. Differentially methylated bins were extracted with a SLIM-adjusted P value threshold of <0.01 and a >25% change in methylation level.

Relative methylation levels in each bin for the sorted clones and their parental samples clones were calculated using the pyBigWig library (v.0.3.22) (https://github.com/deeptools/pyBigWig) on Python (v.3.10.0), and the resulting BigWig files were visualized in IGV (v.2.16.2)⁸⁰.

Experiments using yeast

Strains

S. cerevisiae BY4741 (MATa his3∆0 leu2∆0 met15∆0 ura3∆0) was used for the yeast CloneSelect experiments.

Barcode plasmid pool preparation

To generate the yeast CloneSelect barcode library, a semi-random oligonucleotide pool, KI#200, encoding 5′-CCGWSNSWSNSWSNSWSNSNGTG-3′, was chemically synthesized (Supplementary Table 2) and amplified by PCR in a 40 µl reaction containing 2 µl of 0.01 µM template, 2 µl each of 10 µM forward primer SI#368 and 10 µM reverse primer SI#369, 0.8 µl of Phusion High-fidelity DNA Polymerase (NEB, no. M0530S), 8 µl of 5× Phusion HF Buffer (NEB, no. B0518S) and 2 µl of 2 mM dNTPs. The thermal cycling conditions were as follows: 98 °C for 30 s; 35 cycles of 98 °C for 10 s, 68 °C for 20 s and 72 °C for 5 s; followed by a final extension at 72 °C for 5 min. The PCR product was analyzed on a 2% agarose gel, size-selected and purified using the FastGene PCR/Gel Extraction Kit (Nippon Genetics, no. FG-91302).

The purified barcode fragment was then assembled into the cloning backbone plasmid pKI110 by Golden Gate Assembly using BsmBI (NEB, no. R0580S). Two assembly reactions were performed, each in a 25 µl volume containing 500 fmol barcode fragments, 50 fmol backbone plasmid, 0.5 µl of BsmBI (NEB, no. R0580S), 0.5 µl of T4 DNA Ligase (Nippon Gene, no. 317-00406), 2.5 µl of 10× T4 DNA Ligation Reaction Buffer (NEB, no. B0202S) and 0.125 µl of 60 mg ml⁻¹ BSA (NEB, no. B9001S). The thermal cycling conditions were as follows: 15 cycles of 37 °C for 5 min and 20 °C for 5 min, followed by 55 °C for 30 min for complete backbone digestion.

For bacterial transformation, 5 µl of the assembly product was used to transform 50 µl of DH5α chemically competent cells (NEB, no. C2987I) following the manufacturer’s high-efficiency transformation protocol. After a 1 h outgrowth in 1 ml of SOC medium (NEB, no. B9020S) at 37 °C, cells were plated on four LB agar plates containing 100 µg ml⁻¹ ampicillin (Wako, no. 014-23302). Diluted samples were also plated to estimate clone complexity. Random clones were isolated and analyzed by genotyping PCR using primers KI#169 and KI#170 to validate the presence of the expected barcode insert.

To construct the Pool-100 plasmid pool, 100 colonies were isolated, dissolved in 80 µl of LB medium containing 100 µg ml⁻¹ ampicillin, combined in 5 µl aliquots and cultured overnight at 37 °C. Plasmid DNA was extracted using the FastGene Plasmid Mini Kit (Nippon Genetics, no. FG-90502). The Pool-1580 was constructed by scraping colonies from a plate with ~1,000 colony-forming units into 1.5 ml LB medium with 100 µg ml⁻¹ ampicillin. Cells were centrifuged at 15,000g for 2 min, and the supernatant was discarded. Plasmid DNA pools were then purified from the collected cells using the FastGene PCR/Gel Extraction Kit (Nippon Genetics, no. FG-91302).

Barcoding of cells and introduction of genome editing reagents

For barcoding cells and introducing genome editing reagents, we used the Frozen-EZ Yeast Transformation II kit (Zymo Research, no. T2001) with slight modifications. Cells were initially pre-cultured in 5 ml of YPDA or SC–Dropout medium (adjusted to meet auxotrophic requirements for plasmid maintenance) in a cell culture tube rotating overnight at 30 °C. The following day, cells were cultured in 5 ml of fresh YPDA medium with a starting optical density at 600 nm (OD₆₀₀) of 0.3 and incubated until the OD₆₀₀ reached 0.8–1.0. After preparing competent cells according to the manufacturer’s protocol, plasmid DNA and 50 µl of competent cells were added to a 1.5 ml tube, mixed thoroughly with 500 µl of EZ3 solution as per the manufacturer’s protocol and incubated at 30 °C for 1 h with rotation. The cell sample was then centrifuged at 15,000g for 2 min, and the supernatant was discarded. For recovery, 2.5 ml of YPDA medium was added, and cells were allowed a 2 h outgrowth at 30 °C with rotation. After recovery, cells were centrifuged, the medium was removed and the cells were washed twice with 1 ml of TE buffer. Finally, cells were spread on SC–Dropout agar plates and incubated for 2–4 days at 30 °C.

Barcoding of cells

When the background BY4741 cells were transformed with the barcode plasmid library containing the HIS3 marker, YPDA medium was used for pre-culturing, and SC–His+Ade plates were used for selecting transformants. For pooled cell barcoding, the reaction was scaled up to transform 250 µl of competent cells using 200 ng of plasmid DNA. Colonies that formed on selective plates were pooled and collected by scraping with 3–4 ml of SC–His+Ade medium. For barcoding cells with a single barcode plasmid clone, 200 ng of plasmid was used to transform 15 µl of competent cells.

Introduction of genome editing reagents

When cells containing the barcode plasmid with the HIS3 marker were subjected to clone isolation, they underwent two rounds of transformation: first with the constitutively active Target-AID plasmid pKI086 containing the LEU2 marker and then with the targeting gRNA expression plasmid containing the URA3 marker. For the first transformation, cells were pre-cultured in SC–His+Ade medium and selected on SC–His–Leu+Ade plates. For the second transformation, cells were pre-cultured in SC–His–Leu+Ade medium and selected on SC–His–Leu–Ura+Ade plates.

When transforming the background BY4741 cells with one of the galactose-inducible Cas9-based enzyme plasmids (Cas9, dCas9, dCas9-PmCDA1, dCas9-PmCDA1-UGI, nCas9, nCas9-PmCDA1 or nCas9-PmCDA1-UGI) containing the LEU2 marker along with a CAN1-targeting gRNA plasmid containing the URA3 marker, YPDA medium was used for pre-culturing, and transformants were selected on SC–Leu–Ura+Ade plates.

For barcode-specific reporter activation within a complex barcoded population, the reaction was scaled up to transform 250 µl of competent cells with 200 ng of the enzyme plasmid and 200 ng of the targeting gRNA plasmid. For smaller-scale transformations, 200 ng of the enzyme plasmid and 200 ng of the target gRNA plasmid were used to transform 15 µl of competent cells.

Barcode sequencing library preparation

The barcode sequencing libraries of the plasmid DNA pools were prepared using a two-step PCR method. The first-round PCR was performed in a 40 µl volume, containing 1.0 µg of template DNA, 1 µl each of 10 µM forward primer KI#169 and 10 µM reverse primer KI#289, 0.4 µl of Phusion High-Fidelity DNA Polymerase (NEB, no. M0530S), 8 µl of Phusion HF Buffer (NEB, no. B0518S) and 0.8 µl of 10 mM dNTPs (Takara, no. 4030). The thermal cycling conditions were as follows: 98 °C for 30 s; 20 cycles of 98 °C for 10 s, 61 °C for 20 s and 72 °C for 25 s; with a final extension at 72 °C for 5 min. Each PCR product was size-selected on a 2% agarose gel, purified and eluted into 50 µl of ddH₂O using the PCR/Gel Extraction Kit (Nippon Genetics, no. FG-91302).

To add Illumina sequencing adaptors and custom indices, a second-round PCR was performed in a 40 µl volume containing 2 µl of the first-round product, 1 µl each of 10 µM P5 and P7 custom index primers, 0.4 µl of Phusion High-Fidelity DNA Polymerase (NEB, no. M0530S), 8 µl of 5× Phusion HF Buffer (NEB, no. B0518S) and 0.8 µl of 10 mM dNTPs (Takara, no. 4030). The thermal cycling conditions were as follows: 98 °C for 30 s; 15 cycles of 98 °C for 10 s, 60 °C for 10 s and 72 °C for 1 min; followed by a final extension at 72 °C for 5 min. Custom indices for the second-round PCR products are listed in Supplementary Table 3. Each second-round PCR product was size-selected and purified using the PCR/Gel Extraction Kit (Nippon Genetics, no. FG-91302).

The sequencing libraries were pooled, quantified by qPCR with the Kapa Library Quantification Kit for Illumina (Kapa Biosystems, no. KK4824), combined in equimolar ratios and analyzed by paired-end sequencing using Illumina HiSeq 2500.

To identify barcodes in the yeast CloneSelect plasmids introduced into cells, yeast cells were first centrifuged at 20,000g for 3 min, and the supernatant was discarded. The cell pellet was resuspended in 20 µl of Zymolyase Buffer containing 2.5 mg ml⁻¹ Zymolyase (Zymo Research, no. E1005) and 500 µl of Solution I Buffer (supplied with Zymolyase, no. E1005) containing 0.1 M EDTA and 1 M sorbitol. The sample was incubated at 37 °C for 1 h, centrifuged at 20,000g for 1 min, and the supernatant was discarded. The cell lysate was then treated with 250 µl of Solution II Buffer (supplied with Zymolyase, no. E1005) containing 20 mM EDTA, 50 mM Tris-HCl and 1% SDS and then incubated at 65 °C for 30 min. After this, 100 µl of 5 M potassium acetate was added, and the sample was incubated on ice for 30 min, followed by centrifugation at 20,000g for 3 min. The supernatant was transferred to a 1.5 ml tube, and plasmid DNA was precipitated by adding 400 µl of isopropanol, followed by a cleanup with 400 µl of 70% ethanol. The DNA pellet was resuspended in 50 µl of ddH₂O containing 10 µg ml⁻¹ RNase and incubated at 65 °C for 10 min. The sequencing library for each sample was prepared using the same method described above for the plasmid DNA pools, with custom indices for the second-round PCR detailed in Supplementary Table 3.

Analysis of reporter activation efficiency

To evaluate the efficiency of gRNA-dependent, barcode-specific mCherry reporter activation, we treated three independent barcoded cell samples with their corresponding gRNAs in a 3 × 3 assay. Each sample was spread on SC–His–Leu–Ura+Ade agar plates, scraped, inoculated into a 1.5 ml tube containing 500 µl of SC–His–Leu–Ura+Ade medium and cultured for 2–4 days at 30 °C.

A 20 µl aliquot of each pre-cultured sample was mixed with 180 µl of SC–His–Leu–Ura+Ade medium and transferred to a flat-bottom transparent 96-well plate (Greiner Bio-One, no. 655090). mCherry fluorescence intensities, normalized by OD₅₉₅ values, were measured using the Infinite 200 PRO plate reader (TECAN) with TECAN i-control software (v.1.10.4.0). For microscopic observations, 2.5 µl of each cell sample was placed on a glass slide, covered with a coverslip and observed under a BZ-X710 microscope (Keyence) with ×20 and ×40 objective lenses.

To directly measure the GTG→ATG conversion rate in each sample, cells were collected from selective plates and lysed with DNAZol (COSMO BIO, no. DN127) according to the manufacturer’s protocol. Sequencing libraries were prepared using a two-step PCR method. The first-round PCR was conducted in a 32 µl reaction containing 1.6 µl of cell lysate, 1.6 µl each of 10 µM forward primer KI#168 and 10 µM reverse primer KI#169, 0.64 µl of Phusion High-Fidelity DNA Polymerase (NEB, no. M0530S), 6.4 µl of Phusion HF Buffer (NEB, no. B0518S) and 0.64 µl of 10 mM dNTPs (Takara, no. 4030). Thermal cycling conditions were as follows: 98 °C for 30 s; 30 cycles of 98 °C for 10 s, 61 °C for 10 s and 72 °C for 1 min; with a final extension at 72 °C for 5 min. The remaining library preparation and sequencing followed the same protocols described for barcode sequencing, with custom indices for the second-round PCR detailed in Supplementary Table 3.

Isolation and analysis of barcoded colonies

After barcode-specific reporter activation in a complex population, cells from test and control conditions were spread on SC–His–Leu–Ura+Ade agar plates and imaged under a blue light illuminator (FAS-IV, Nippon Genetics) to isolate mCherry⁺ or mCherry⁻ colonies. Colonies were then isolated into 96-well cell culture plate wells containing 98 µl of SC–His–Leu–Ura+Ade medium and cultured overnight at 30 °C.

For analysis, samples were measured for mCherry fluorescence intensities normalized by OD₅₉₅ values using an Infinite 200 PRO plate reader (TECAN) with TECAN i-control software (v.1.10.4.0). The same isolated colonies were also subjected to Sanger sequencing to identify their barcode sequences and assess base editing outcomes. Barcode DNA fragments were obtained using the same protocols for cell lysis, first-round PCR and PCR cleanup as in the reporter activation efficiency analysis. Each PCR product was analyzed by Sanger sequencing with sequencing primer SI#658, and sequencing traces were processed using PySanger (https://github.com/ponnhide/PySanger).

Canavanine assay

Genome editing efficiencies of different Cas9-based genome editing enzymes (Cas9, dCas9, dCas9-PmCDA1, dCas9-PmCDA1-UGI, nCas9, nCas9-PmCDA1 and nCas9-PmCDA1-UGI) were estimated using a canavanine assay. In this assay, Cas9-based enzymes under a galactose-inducible GAL1/10 promoter were introduced to cells along with a gRNA targeting the arginine transporter gene CAN1, allowing assessment of knockout efficiency through cell survival in the presence of the toxic arginine analogue canavanine.

To induce genome editing, cells containing both enzyme and gRNA plasmids were first cultured in SC–Leu–Ura medium with 2% glucose at 30 °C until saturation. Cells were then resuspended in SC–Leu–Ura medium with 2% raffinose at a 16-fold dilution and cultured at 30 °C until saturation. Finally, cells were resuspended in SC–Leu–Ura medium containing 2% raffinose and 0.02% galactose at a 32-fold dilution and cultured at 30 °C for 2 days.

Each sample was spread on SC–Leu–Ura–Arg+Ade plates and SC–Leu–Ura–Arg+Ade plates containing 60 mg ml⁻¹ canavanine. Plates were incubated at 30 °C for 2–4 days to estimate colony-forming units and for spot assays. After examining colony-forming units, colonies were scraped from the SC–Leu–Ura–Arg+Ade control plates for genomic DNA extraction to assess mutation spectra by high-throughput sequencing.

For DNA extraction, 20 µl of cells at OD₅₉₅ of 1.0 were lysed in 100 µl of DNAzol (COSMO BIO, no. DN127) and incubated at room temperature for 15 min. The lysate was mixed with 30 µl of 1 M NaCl and 50 µl of 100% ethanol, then centrifuged at 15,000g for 10 min. The supernatant was discarded, and the pellet was washed with 550 µl of 70% ethanol. After air-drying, the sample was resuspended in 50 µl of ddH₂O.

Amplicon sequencing libraries were prepared for each sample using a two-step PCR method in triplicate. The first-round PCR was conducted in a 40 µl volume containing 2 µl of template DNA, 2 µl each of 10 µM forward primer no. KN85F3 and 10 µM reverse primer no. KN85R2, 0.8 µl of Phusion High-Fidelity DNA Polymerase (NEB, no. M0530S), 8 µl of Phusion HF Buffer (NEB, no. B0518S) and 0.8 µl of 10 mM dNTPs (Takara, no. 4030). The thermal cycling conditions were as follows: 98 °C for 30 s; 30 cycles of 98 °C for 10 s, 60 °C for 10 s and 72 °C for 1 min; with a final extension at 72 °C for 5 min. Control samples were prepared using primer pair HO2F2–HO2R2. The remaining library preparation and sequencing followed the same protocols described for barcode sequencing, with custom indices for the second-round PCR detailed in Supplementary Table 3.

Mutational spectra analysis

Amplicon sequencing reads obtained to assess mutational patterns at the CAN1 target site, induced by each Cas9-based genome editing enzyme, were processed using a previously established pipeline³⁸. The codes specific to this analysis are available at https://github.com/yachielab/CloneSelect_v1/tree/main/Mutational_Spectra_Analysis.

Experiments using E.
coli

Preparation of cells for various Bacterial CloneSelect systems

Cell samples containing single barcode plasmids were prepared for different Bacterial CloneSelect systems (Supplementary Table 2). For the EGFP reporter-based system, the plasmid was introduced into BL21(DE3) E. coli cells (NEB, no. C2527I). For the blasticidin and Zeocin resistance marker-based systems, plasmids were introduced into T7 Express chemically competent E. coli cells (NEB, no. C2566I), following the manufacturer’s high-efficiency transformation protocols. Transformants were selected on LB agar plates containing 100 µg ml⁻¹ ampicillin (Wako, no. 014-23302) and/or 50 µg ml⁻¹ kanamycin (Wako, no. 111-00344).

Barcode plasmid pool preparation

To generate the bacterial CloneSelect barcode library for the Zeocin resistance marker, a semi-random oligonucleotide pool KI#405 encoding 5′-ATGCCGVNNVNNVNNVNNVNNTAA-3′ was chemically synthesized (Supplementary Table 2). This sequence includes a start codon (ATG), the antisense strand of the 5′-CGG-3′ PAM, a quintuple repeat of VNN (V = non-T) and a stop codon (TAA). The VNN repeat restricts the appearance of in-frame stop codons upstream of the reporter.

The oligonucleotide pool was amplified by PCR in a 20 µl reaction containing 1 µl of 1 µM template, 1 µl each of 10 µM forward primer SI#368 and 10 µM reverse primer SI#369, 0.4 µl of Phusion High-Fidelity DNA Polymerase (NEB, no. M0530L), 4 µl of 5× Phusion HF Buffer (NEB, no. B0518S) and 0.4 µl of 10 mM dNTPs. The thermal cycling conditions were as follows: 98 °C for 30 s; 20 cycles of 98 °C for 10 s, 68 °C for 20 s and 72 °C for 20 s; followed by a final extension at 72 °C for 5 min. The PCR product was analyzed on a 2% agarose gel, size-selected and purified using the FastGene PCR/Gel Extraction Kit (Nippon Genetics, no. FG-91302).

The purified barcode fragment was assembled into the cloning backbone plasmid pKI243 by Golden Gate Assembly using BsmBI. The assembly reaction was performed in a 12.5 µl volume containing 2.91 fmol barcode fragments, 14.9 fmol backbone plasmid, 0.25 µl of BsmBI (NEB, no. R0580L), 0.5 µl of T4 DNA Ligase (Nippon Gene, no. 317-00406), 1.25 µl of 10× T4 DNA Ligation Reaction Buffer (NEB, no. B0202S) and 0.62 µl of 2 mg ml⁻¹ BSA (NEB, no. B9001S). Thermal cycling conditions were 15 cycles of 37 °C for 5 min and 20 °C for 5 min, followed by 55 °C for 30 min for complete backbone digestion.

For transformation, 3 µl of the assembly product was used to transform 65 µl of T7 Express chemically competent cells (NEB, no. C2566I) following the high-efficiency transformation protocol. After a 1 h outgrowth in 500 µl of SOC medium (NEB, no. B9020S) at 37 °C, the cell sample was plated in 250 µl portions on three LB agar plates containing 100 µg ml⁻¹ ampicillin (Wako, no. 014-23302). Diluted samples were also plated on selective plates to estimate clone complexity. Assembly quality and efficiency were checked by isolating 12 random clones and validating the barcode inserts by Sanger sequencing, with 11 out of 12 clones showing the expected barcode insert.

To construct the Pool-100 plasmid pool, 100 colonies were isolated, each resuspended in 80 µl of LB medium with 100 µg ml⁻¹ ampicillin, combined in 5 µl aliquots and cultured overnight at 37 °C. Plasmid DNA was extracted using the FastGene Plasmid Mini Kit (Nippon Genetics, no. FG-90502). The Pool-1550 was constructed by scraping colonies from a plate with ~1,000 colony-forming units into 1.5 ml LB medium with 100 µg ml⁻¹ ampicillin. The barcode plasmid libraries were used to transform T7 Express chemically competent cells (NEB, no. C2566I) to establish barcoded E. coli cell populations.

Barcode sequencing library preparation

For the Pool-100 and Pool-1550 barcode plasmid libraries, barcode sequencing libraries were prepared in triplicate using a two-step PCR method. The first-round PCR was performed in five separate 40 µl reactions, each containing 2.0 ng of plasmid template DNA, 1 µl each of 10 µM forward primer KI#403 and 10 µM reverse primer KI#404, 0.4 µl of Phusion High-Fidelity DNA Polymerase (NEB, no. M0530L), 8 µl of Phusion HF Buffer (NEB, no. B0518S) and 0.8 µl of 10 mM dNTPs (Takara, no. 4030). Thermal cycling conditions were as follows: 98 °C for 30 s; 20 cycles of 98 °C for 10 s, 54 °C for 20 s and 72 °C for 25 s; followed by a final extension at 72 °C for 5 min. For each replicate, the five PCR products were pooled, size-selected on a 2% agarose gel, purified and eluted in 30 µl of ddH₂O using the PCR/Gel Extraction Kit (Nippon Genetics, no. FG-91302).

In the second-round PCR, Illumina sequencing adaptors and custom indices were added to each first-round PCR product. Each 40 µl reaction contained 2 µl of the first-round PCR product, 1 µl each of 10 µM P5 and P7 custom index primers, 0.4 µl of Phusion High-Fidelity DNA Polymerase (NEB, no. M0530L), 8 µl of 5× Phusion HF Buffer (NEB, no. B0518S) and 0.8 µl of 10 mM dNTPs (Takara, no. 4030). The thermal cycling conditions were as follows: 98 °C for 30 s; 15 cycles of 98 °C for 10 s, 60 °C for 10 s and 72 °C for 60 s; followed by a final extension at 72 °C for 5 min. Custom indices for the second-round PCR products are listed in Supplementary Table 3.

The second-round PCR products were size-selected and purified using the PCR/Gel Extraction Kit (Nippon Genetics, no. FG-91302). The sequencing libraries were pooled, quantified by qPCR using the Kapa Library Quantification Kit for Illumina (Kapa Biosystems, no. KK4824), combined in equimolar ratios and analyzed by paired-end sequencing using Illumina MiSeq.

Introduction of genome editing reagents

To introduce a plasmid containing ABE-7.10 and a gRNA to barcoded cells, we used the Mix&Go! E. coli Transformation Kit (Zymo Research, no. T3001) following the manufacturer’s protocol. Transformants were selected by plating the transformation reaction on LB agar plates containing 100 µg ml⁻¹ ampicillin (Wako, no. 014-23302) and 50 µg ml⁻¹ kanamycin (Wako, no. 111-00344) and incubating overnight at 37 °C.

For experiments involving induction with Ara (Sigma-Aldrich, no. A3256-10MG) and IPTG (ThermoFisher Scientific, no. 15529019), cells were cultured overnight at 37 °C in medium containing 100 mM Ara and 0.1 mM IPTG before analysis. For barcoded cell isolation using the Zeocin resistance marker-based system, a low-salt LB medium adjusted to pH 7.5 with 1 M NaOH (Nakalai, no. 37421-05) was used to optimize Zeocin activity.

For genome editing and selection of reporter-activated cells without inducers, we used 100 µg ml⁻¹ Zeocin (Invitrogen, no. R25001) or 100 µg ml⁻¹ blasticidin S (Wako, no. 029-18701), as the leaky expression from inducible promoters in the absence of inducers was sufficient for gene editing while maintaining high cell viability. Details of the genome editing plasmids used in this study are provided in Supplementary Table 2.

Analysis of reporter activation efficiency

To evaluate the efficiency of gRNA-dependent, barcode-specific EGFP reporter activation, 200 µl of cell samples were transferred into a flat-bottom transparent 96-well plate (Greiner Bio-One, no. 655090) and analyzed using the Infinite 200 PRO plate reader (TECAN) with TECAN i-control software (v.1.10.4.0) to measure EGFP fluorescence intensities normalized to OD₅₉₅ values.

For microscopic observation, 2.5 µl of each cell sample was placed on a glass slide (MATSUNAMI, no. S2441), gently covered with a glass coverslip and examined under a BZ-X710 microscope (Keyence) using ×20 and ×40 objective lenses.

Isolation and analysis of barcoded colonies

After barcode-specific activation of the Zeocin resistance marker in the barcoded cell population, barcodes from colonies under test and control conditions were analyzed by Sanger sequencing. For each colony, the barcode region was amplified by PCR in a 20 µl reaction containing 1 µl of cell suspension, 0.5 µl each of 10 µM forward primer KI#403 and 10 µM reverse primer KI#404, 0.2 µl of Phusion High-Fidelity DNA Polymerase (NEB, no. M0530L), 4 µl of Phusion HF Buffer (NEB, no. B0518S) and 0.4 µl of 10 mM dNTPs (Takara, no. 4030). The thermal cycling conditions were as follows: 98 °C for 30 s; 30 cycles of 98 °C for 10 s, 54 °C for 20 s and 72 °C for 30 s; followed by a final extension at 72 °C for 5 min.

The PCR products were analyzed on a 2% agarose gel and transferred to wells of a 96-well PCR plate for cleanup using 20 µl of AMPure XP beads (Beckman Coulter, no. A63881) according to the manufacturer’s protocol. Sanger sequencing was conducted using primer KI#403, and sequencing traces were analyzed with PySanger (https://github.com/ponnhide/PySanger).

Reporting summary

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