ONE-STEP tagging: a versatile method for rapid site-specific integration by simultaneous reagent delivery

Abstract

We present a novel, versatile genome editing method termed ONE-STEP tagging, which combines CRISPR-Cas9-mediated targeting with Bxb1 integrase-based site-specific integration for efficient, precise, and scalable protein tagging. Applied in human-induced pluripotent stem cells (hiPSCs), cancer cells and primary T cells, this system enables rapid generation of endogenously tagged proteins. By enhancing the nuclear localization signal of the catalytically superior eeBxb1 integrase and co-delivering a DNA-PK inhibitor, we achieved up to ∼90% integration efficiency at the ACTR10 locus in hiPSCs. ONE-STEP tagging is robust across loci and cell types and supports large DNA cargo integration, with efficiencies reaching 16.6% for a 14.4 kb construct. The method also enables multiplexed tagging of multiple proteins within the same cell and simultaneous CRISPR-based editing at secondary loci, such as gene knockouts or homology-directed repair. Importantly, we demonstrate successful application in primary T cells by targeting the T cell receptor locus while simultaneously knocking out B2M, a key step towards generating immune-evasive, off-the-shelf chimeric antigen receptor T cells. Additionally, we introduce a dual-cassette version of the method compatible with universal donor plasmids, allowing use of entirely off-the-shelf reagents. Together, these advances establish ONE-STEP tagging as a powerful tool for both basic and therapeutic genome engineering.

Graphical Abstract

Figure 1.

Schematic of the ONE-STEP tagging technology. The system involves insertion of attachment site (attP) via CRISPR–Cas9 mediated by HDR and an ssODN as template. Upon successful HDR, the Bxb1 integrase results in site-specific integration of the mNeonGreen cargo. All reagents are delivered simultaneously. Created in BioRender. Bassett, A. (2025) https://BioRender.com/lwtgerr.

Figure 2.

Optimization of ONE-STEP tagging for efficient cargo integration. (A) Tagging efficiency at the ACTR10 locus in hiPSCs using different concentrations of circularized mNeonGreen donor plasmid (1-, 3-, or 6-fold molar excess relative to Bxb1), with a constant 500 ng of Bxb1 plasmid. The “Tagging efficiency” condition includes all required components; the “Random integration control” omits the ssODN–attP (NA), resulting in <1% non-specific integration. A 6-fold excess of mNeonGreen donor achieved ∼10% tagging efficiency (n = 3). Error bars represent standard deviation. Data were analysed using two-sided t test. (B). Comparison of tagging efficiency at the ACTR10 locus in hiPSCs using different Bxb1 variants and NLS: WT Bxb1 with C-terminal SV40 NLS, WT Bxb1 with C-terminal 2 × NPL NLS, and engineered eeBxb1 with C-terminal 2 × NPL NLS. The eeBxb1 2 × NPL NLS variant achieved ∼25% tagging efficiency with a 6-fold excess of donor (n = 3). Error bars represent standard deviation. Data were analysed using two-sided t test. (C). Effect of DNA-PK inhibitors on tagging efficiency. Cells were treated immediately after nucleofection, with media replacement after 24 h. Compounds tested: NU7441 (4 μM), IDT HDR Enhancer v2 (1 μM), M3814 (4 μM), and AZD-7648 (0.5 μM). AZD-7648 (0.5 μM) treatment resulted in >80% tagging efficiency at the ACTR10 locus (n = 3). Error bars represent standard deviation. Data were analysed using two-sided t test. (D) Representative flow cytometry plots showing ACTR10-mNeonGreen tagging efficiency in untreated control cells (top) and cells treated with AZD-7648 (bottom). (E) Tagging efficiency at the ACTR10 locus in hiPSCs using a 6-fold molar excess of mNeonGreen donor with different Bxb1 variants/NLS combinations, with or without AZD-7648. The effect of AZD-7648 was variant-dependent, with the largest enhancement observed for WT Bxb1 with SV40 NLS. Error bars represent standard deviation. Data were analysed using two-sided t test. (F) Representative flow cytometry plot illustrating 14.4 kb integration at the BFP reporter gene in hiPSCs. Successfully tagged cells exhibited loss of BFP and gain of RFP signal. (G) Bar chart showing integration of a 14.4 kb cargo into the BFP reporter gene integrated as a single copy at the ROSA26 locus in the kolf_2_C1 hiPSC line. ONE-STEP tagging achieved ∼16% efficiency, compared to ∼1% using eePASSIGE (n = 3). Error bars represent standard deviation. Data were analysed using two-sided t test. (H) Benchmarking of ONE-STEP tagging against HDR and eePASSIGE at the ACTR10 locus in hiPSCs. mNeonGreen integration efficiency was ∼30% with HDR (n = 2), ∼0.2% with eePASSIGE (n = 3), and ∼80% using ONE-STEP tagging (n = 3). Error bars represent standard deviation. Data were analysed using two-sided t test.

Figure 3.

Efficient and specific integration using heterotypic recombination sites enables multiplexed tagging. (A) Schematic representation of attP/attB-GA and attP/attB-GT cassettes. Created in BioRender. Bassett, A. (2025) https://BioRender.com/lwtgerr. (B) Evaluation of mNeonGreen integration at the ACTR10 locus using attP-GA or attP-GT variants and matched or mismatched attB/attP dinucleotide combinations in A1ATD hiPSC cells. Specific pairing of attB and attP dinucleotides (GA–GA or GT–GT) results in higher integration efficiency compared to mismatched pairs (n = 3). Error bars represent standard deviation. Data were analysed using two-sided t test. (C) Schematic of the multiplexed protein tagging strategy. (D) Representative flow cytometry plot showing integration efficiencies at the ACTR10 locus (mNeonGreen), MAP4 locus (mCherry), and simultaneous integration at both loci (double-positive for mNeonGreen and mCherry) in A1ATD hiPSCs. (E) Stacked bar chart showing quantification of flow cytometry data from n = 3 biological replicates. Error bars represent standard deviation. A substantial population of double-positive cells (up to ∼26%) was observed.

Figure 4.

Simultaneous integration and genome editing using ONE-STEP tagging. (A) Schematic of experimental design. Created in BioRender. Bassett, A. (2025) https://BioRender.com/lwtgerr. Reagents for tagging MAP4 with mCherry were co-delivered with a sgRNA and HDR template to convert BFP to GFP in a BFP reporter cell line (BFP reporter kolf_2_C1 hiPSCs). The mCherry signal indicates MAP4 tagging, while GFP gain monitored HDR at the BFP locus. (B) Representative flow cytometry plot showing integration at the MAP4 locus (mCherry) and simultaneous HDR at the BFP locus (GFP); ∼8% double-positive cells were detected. (C) Representative flow cytometry plot showing that ∼80% of mCherry-positive cells exhibited GFP conversion, indicating successful HDR. (D) Bar chart quantifying HDR events in mCherry-tagged versus untagged cells. Approximately 80% of mCherry-positive cells showed HDR, compared to ∼30% of untagged cells. Data represent n = 3 biological experiments; error bars indicate standard deviation. Data were analysed using two-sided t test. (E). Schematic of experimental design for simultaneous tagging of ACTR10 with mNeonGreen and KO of the BFP gene in BFP reporter kolf_2_C1 hiPSCs. mNeonGreen-tagging reagents for ACTR10 were co-delivered with a BFP sgRNA into the BFP reporter line (BFP reporter Kolf_2_C1). Successful KO events were monitored by loss of BFP expression. Created in BioRender. Bassett, A. (2025) https://BioRender.com/lwtgerr. (F) Representative flow cytometry plot showing integration at the ACTR10 locus (mNeonGreen) and simultaneous BFP KO; ∼46% of cells were mNeonGreen-positive and BFP-negative. (G) Representative flow cytometry plot showing that ∼90% of ACTR10-mNeonGreen-positive cells exhibited BFP loss. (H) Bar chart quantifying KO events in mNeonGreen-tagged versus untagged cells. Approximately 90% of mNeonGreen-positive cells exhibited BFP loss, compared to ∼50% of untagged cells. Data represent n = 3 biological experiments; error bars indicate standard deviation. Data were analysed using two-sided t test.

Figure 5.

Targeted integration at the TCR locus and simultaneous KO of B2M. (A) Schematic of experimental design in primary T cells. An eGFP-expressing 4.4 kb cargo was integrated into the TRAC locus using an ssODN containing an attP site as an HDR template, together with a sgRNA targeting the TRAC locus. A B2M-targeting guide RNA can be co-delivered to achieve simultaneous KO at the B2M locus. Created in BioRender. Bassett, A. (2025) https://BioRender.com/lwtgerr. (B). Left: Bar chart showing ∼12% tagging efficiency at the TRAC locus in primary T cells. Data represent triplicate experiments; error bars indicate standard deviation. Data were analysed using two-sided t test. Right: Representative flow cytometry plot illustrating successful eGFP integration. (C) Schematic and flow cytometry analysis of simultaneous eGFP integration at the TRAC locus and B2M KO in primary T cells. All components were delivered in a single nucleofection. Seven days post-nucleofection, flow cytometry analysis revealed that >90% of eGFP-positive cells (indicating successful TRAC targeting) lacked B2M expression, compared to ∼30% B2M KO in the total cell population. Error bars indicate standard deviation. Data were analysed using two-sided t test.

Figure 6.

Expanding the ONE-STEP platform with dual-cassette integration and multiplexed tagging. (A) Schematic of dual-cassette ONE-STEP tagging strategy. A single-stranded DNA HDR template (∼200 nt) containing two heterologous attP sites (attP-GA and attP-GT) and ∼50 bp homology arms was used in combination with a plasmid donor containing cargo flanked by corresponding attB variants. Created in BioRender. Bassett, A. (2025) https://BioRender.com/lwtgerr. (B) Bar chart comparing mNeonGreen integration efficiency at the ACTR10 locus using the original ONE-STEP technology (∼80%) and the dual-cassette strategy (∼60%) in A1ATD hiPSC cells. Error bars indicate standard deviation. Data were analysed using two-sided t test. (C). Bar chart showing mNeonGreen integration at three additional loci (LMNA, FBL, and MAP4) in the A1ATD hiPSC line. Tagging efficiencies ranged from 2.1% to 10.8%, depending on the locus. Data represent n = 3 biological experiments; error bars indicate standard deviation. Data were analysed using two-sided t test. (D) Fluorescence microscopy images confirming subcellular localization of mNeonGreen consistent with known localization patterns for all four targeted proteins (ACTR10, LMNA, FBL, and MAP4) in the A1ATD hiPSC cell line. (E) Bar chart showing extension of ONE-STEP tagging to additional cell types. Integration at the MAP4 locus resulted in tagging efficiencies of 6.6% in Kolf_2_C1 hiPSC cells and 28.6% in K562 lymphoblast cells. Data represent n = 3 biological experiments; error bars indicate standard deviation. Data were analysed using two-sided t test. (F) Schematic of multiplexed gene integration strategy using dual-cassette donors. ACTR10 was targeted with GA-mNeonGreen-GT, and MAP4 with CT-mCherry-AG, in a single nucleofection. Created in BioRender. Bassett, A. (2025) https://BioRender.com/lwtgerr. (G) Flow cytometry analysis demonstrating successful dual tagging, with ∼6% of cells expressing both mNeonGreen and mCherry in A1ATD hiPSC cell line. (H) Bar chart quantifying dual-tagging efficiency across triplicate experiments; error bars indicate standard deviation. (I) Double-positive and single-positive cell populations were sorted by flow cytometry, and site-specific integration at ACTR10 and MAP4 loci was validated by ddPCR using junction-specific probes for mNeonGreen and mCherry. Error bars indicate standard deviation. Data were analysed using two-sided t test.