Cas12a-knock-in mice for multiplexed genome editing, disease modelling and immune-cell engineering

Abstract

The pleiotropic effects of human disease and the complex nature of gene-interaction networks require knock-in mice allowing for multiplexed gene perturbations. Here we describe a series of knock-in mice with a C57BL/6 background and with the conditional or constitutive expression of LbCas12a or of high-fidelity enhanced AsCas12a, which were inserted at the Rosa26 locus. The constitutive expression of Cas12a in the mice did not lead to discernible pathology and enabled efficient multiplexed genome engineering. We used the mice for the retrovirus-based immune-cell engineering of CD4⁺ and CD8⁺ T cells, B cells and bone-marrow-derived dendritic cells, for autochthonous cancer modelling through the delivery of multiple CRISPR RNAs as a single array using adeno-associated viruses, and for the targeted genome editing of liver tissue using lipid nanoparticles. We also describe a system for simultaneous dual-gene activation and knockout (DAKO). The Cas12a-knock-in mice and the viral and non-viral delivery vehicles provide a versatile toolkit for ex vivo and in vivo applications in genome editing, disease modelling and immune-cell engineering, and for the deconvolution of complex gene interactions.

Fig. 1. Generation of conditional and constitutive LbCas12a and enAsCas12a knock-in mice.

a, Schematic of the LSL-LbCas12a and LSL-enAsCas12a-HF1 Rosa26-targeting constructs for the two conditional transgenic lines. The backbone is Ai9 Rosa26-targeting vector. LbCas12a, labelled with HA tag and eGFP, is expressed under a CAG promoter. LoxP-(Stop)3XPolyA-LoxP (LSL) allows Cre-dependent conditional expression of LbCas12a protein. LSL-enAsCas12a mice follow the similar design, but are labeled with Myc Tag and use Egl-13 NLS instead of SV40 NLS on the N-terminus. Partial element created in BioRender. Chen, S. (2025) https://BioRender.com/y08x327. b, Generation of the constitutively active LbCas12a and enAsCas12a transgenic mouse line by crossing the corresponding conditional mouse line with the CMV-Cre mouse line. The LSL cassette was excised by Cre, leading to constitutive expression of LbCas12a or enAsCas12a protein. Partial element created in BioRender. Chen, S. (2025) https://BioRender.com/b54g776. c, Widefield fluorescence microscopy illustrating the expression of enAsCas12a-HF1-eGFP protein only in constitutive enAsCas12a mouse, but not in conditional LSL-enAsCas12a mouse or parental C57BL/6 mouse. d, Western blot showing the expression of enAsCas12a-HF1-MycTag protein in enAsCas12a mouse. enAsCas12a-HF1-MycTag protein was not detected in the protein lysate from LSL-enAsCas12a-HF1 and C57BL/6 mouse. Anti-MycTag antibody was used to detect enAsCas12a-HF1 protein and GAPDH was used as the internal control. e, IVIS spectrum imaging of GFP radiance from major organs in LSL-enAsCas12a mice, enAsCas12a-heterozygote mice and enAsCas12a-homozygote mice. Major organs include liver, spleen, kidney, heart, lung and brain. f, Quantification of IVIS GFP radiance. Two-way analysis of variance (ANOVA) with Dunnett’s multiple comparisons test was used to assess significance. For all groups, N = 3 biological replicates. The dash line represents the average background expression from the LSL-enAsCas12a control mice (2.64 × 10⁹). g, Left: white blood cell (WBC) count for constitutive enAsCas12a mice, LbCas12a mice and C57BL/6 control mice. Middle left: red blood cell (RBC) count for constitutive enAsCas12a mice, LbCas12a mice and C57BL/6 control mice. Middle right and right: comparison of lymphocytes/monocytes differential between constitutive enAsCas12a mice, LbCas12a mice and C57BL/6 control mice. One-way ANOVA with Tukey’s multiple comparisons test was used to assess significance. For bar plot, data are shown as mean ± s.e.m. For C57BL/6, N = 7 biological replicates. For enAsCas12a, N = 6 biological replicates. For LbCas12a, N = 6 biological replicates. Exact P values are labelled. h, Quantification of splenic immune cell percentage by flow cytometry. Two-way ANOVA with Tukey’s multiple comparisons test was used to assess significance. All bar plots shown as mean ± s.e.m. Exact P values are labelled. Source data

Fig. 2. Multiplexed immune cell gene editing with LbCas12a mice and enAsCas12a mice.

a, Schematic showing the ex vivo workflow for multiplexed gene editing in primary immune cells (of both LbCas12a mice and LSL-enAsCas12a mice). CD8⁺ T cells and BMDCs were isolated from the spleen and bone marrow of LbCas12a or LSL-enAsCas12a mice, followed by ex vivo culture and retroviral infection. FACS was used to analyse the efficiency of single-cell-level DKO and sort for infected cells for downstream molecular analysis. Created in BioRender. Chen, S. (2025) https://BioRender.com/d80g206. b, Quantification of CD24 or CD11c negative BMDCs percentage for the guides compared with the vector control. Unpaired two-sided t-test was used to assess significance. N = 15 technical replicates for all groups. c, Quantification showing the percentage of CD8a or Thy1 negative CD8⁺ T cells for the guides compared with the vector control. Unpaired two-sided t-test was used to assess significance. N = 3 technical replicates for CD8 groups, and N = 10 technical replicates for Thy1 groups. d, Quantification showing the percentage of Thy1, Cxcr4 and CD4 negative CD4⁺ T cells for the guides compared with the vector control. Unpaired two-sided t-test was used to assess significance. Technical replicates for all groups. N = 10 for Thy1 groups. N = 5 for Cxcr4 groups. N = 3 for CD4 groups. e, Flow cytometry analysis on the CD24-CD43 DKO experiment in LbCas12a BMDCs. Left: representative FACS plot demonstrating CD24 and CD43 expression in BMDCs. Right: quantification of CD24 and CD43 DKO efficiency in different groups. One-way ANOVA with Dunnett’s multiple comparisons test was used to assess significance. N = 3 biological replicates for all groups. f, Flow cytometry analysis on the CD24-CD11c DKO experiment in LbCas12a BMDCs. Left: representative FACS plot demonstrating CD24 and CD11c expression in BMDCs. Right: quantification of CD24 and CD11c DKO efficiency in different groups. One-way ANOVA with Dunnett’s multiple comparisons test was used to assess significance. N = 3 biological replicates for all groups. g, Schematic showing the retroviral vector design for LSL-enAsCas12a immune cell editing. crRNA array was expressed by human U6 promoter (hU6). Cre and mScarlet expression was driven by EFS promoter to induce enAsCas12a-HF1 expression and to label the infected cells, respectively. CD8⁺ T cells and BMDCs were isolated from the spleen and bone marrow of LSL-enAsCas12a mice, as shown in a. Partial element created in BioRender. Chen, S. (2025) https://BioRender.com/b54g776. h, Flow cytometry analysis on the CD24-CD11c DKO experiment in LSL-enAsCas12a BMDCs. Left: representative FACS plots demonstrating CD24 and CD11c expression in BMDCs. Right: quantification of CD24 and CD11c DKO efficiency in different groups. One-way ANOVA with Dunnett’s multiple comparisons test was used to assess significance. N = 4 biological replicates for all groups. i, Targeted sequencing quantifies percent gene modification for the CD24-CD11c DKO experiment in BMDCs. Two-way ANOVA with Dunnett’s multiple comparisons test was used to assess significance. For all groups, N = 3 biological replicates. j, Flow cytometry analysis on the CD80-H2Ab1 DKO experiment in LSL-enAsCas12a BMDCs. Left: representative FACS plots demonstrating CD80 and MHC II expression in BMDCs. Right: quantification of CD80 and MHC II DKO efficiency in different groups. One-way ANOVA with Dunnett’s multiple comparisons test was used to assess significance. N = 3 biological replicates for all groups. k, Targeted sequencing quantifies percent gene modification for the CD80-H2Ab1 DKO experiment in BMDCs. Two-way ANOVA with Dunnett’s multiple comparisons test was used to assess significance. For all groups, N = 3 biological replicates. l, Flow cytometry analysis on the Thy1-CD27 DKO experiment in LSL-enAsCas12a CD8⁺ T cells. Left: representative FACS plots demonstrating CD27 and Thy1 expression in CD8⁺ T cells. Right: quantification of CD27 and Thy1 DKO efficiency in different groups. One-way ANOVA with Dunnett’s multiple comparisons test was used to assess significance. N = 3 biological replicates for all groups. m, Flow cytometry analysis comparing DKO efficiency in BMDCs of LSL-LbCas12a and LSL-enAsCas12a mice. Left: representative FACS plots demonstrating CD24 and CD11c expression in BMDCs. Right: quantification of CD24 and CD11c DKO efficiency in different groups. One-way ANOVA with Tukey’s multiple comparisons test was used to assess significance. N = 4 technical replicates for all groups. All bar plots shown as mean ± s.e.m. Exact P values are labelled. Source data

Fig. 3. In vivo gene editing demonstrated by liver LNP-crRNA targeting and AAV-mediated tumourigenesis.

a, Schematic showing the packaging and delivery of crRNA using LNP to knockout Ttr gene in the liver. Non-targeting control crRNA 1 (NTC1) were packaged as control in the same batch. Constitutive enAsCas12a-HF1 mice were intravenously injected with LNP-crRNA. Serum samples were collected for ELISA to measure and monitor the TTR protein level in serum. Created in BioRender. Chen, S. (2025) https://BioRender.com/j37t723. b, Serum TTR level (µg ml⁻¹) from samples collected at day 0 (before injection), day 6, day 12 and day 20 post injection, via retro-orbital blood draw. The serum TTR level was measured by ELISA. Two independent guides (TTRcr1 and TTRcr2) targeting murine Ttr gene were compared with non-targeting control crRNA (NTC1). Two-way ANOVA with Dunnett’s multiple comparisons test was used to assess significance. Data are shown as mean ± s.e.m. P values are labelled. For all groups, N = 5 biological replicates. c, Pairwise co-occurrence (CO) and mutual exclusivity (ME) analysis for TSGs: TP53, APC, PTEN, RB1, SMAD4 and STK11 on human cancer data from MSK-IMPACT project. Positive log₂(odds ratio) indicated CO, while negative log₂(odds ratio) indicated ME^,. We identified TP53, APC, PTEN and RB1 as a group of co-occur genes out of the six TSGs (surrounded by blue lines). SMAD4 was excluded owing to ME with PTEN, RB1 and STK11. STK11 was excluded because it was mutually exclusive with PTEN and SMAD4. d, Schematic showing the core segment of the AAV construct used for tumour induction, which contained a crRNA expression array (crTSG) that included four guides targeting Trp53, Pten, Apc and Rb1, and a Cre expression cassette for inducing enAsCas12a-HF1 expression. After production and purification, AAV-crTSG and AAV-vector (as the negative control) were either intravenously or intratracheally injected into LSL-enAsCas12a-HF1 mice. Tumour and major organs were isolated for NGS and histology analysis. Partial element created in BioRender. Chen, S. (2025) https://BioRender.com/l72d475. e, Representative IVIS spectrum images detecting GFP signal indicated by total radiance (p s⁻¹ cm⁻² sr⁻¹) for tumour and major organs. Top row being representative AAV-crTSG intravenously injected mouse and the bottom row being representative AAV-vector intravenously injected mouse. f, Representative H&E staining and IHC staining on AAV-induced salivary gland SCC FFPE samples. For IHC, tumour samples were stained with GFP (enAsCas12a-HF1-eGFP) and Ki67 (proliferation marker). Black triangles point out the representative GFP⁺ cancerous cell. For 10× images, scale bars = 300 µm. For 40× images, scale bars = 60 µm. g, Representative H&E staining and IHC staining on AAV-induced lung adenocarcinoma FFPE samples. For IHC, tumour samples were stained with GFP (enAsCas12a-HF1-eGFP). Black triangles point out the representative GFP⁺ cancerous cell. For 10× images, scale bars = 300 µm. For 40× images, scale bars = 60 µm. h, Histology section used for CODEX. Top: global view H&E staining on the AAV-induced salivary gland SCC FFPE sample. Clusters 1, 2, 4 and 10 from unsupervised clustering in the UMAP plot in i were labelled with black box. Scale bar, 1 mm. Bottom: global view GFP staining on the same SCC FFPE sample. Scale bar, 1 mm. i, CODEX results reveal the heterogeneity in the tumour samples. Right: UMAP dimensional reduction plot showing the unsupervised clustering of different cell types in the SCC FFPE sample. Left: spatial display of the 17 cell types overlaying the H&E image. Scale bar, 1 mm. j, Heatmap showing the differential expression of the 22 CODEX markers in the 17 cell-type clusters. The expression level was shown as Z score. Source data

Fig. 4. enAsCas12a knock-in mice demonstrated high on-target and low off-target editing efficiency.

a, Representative allele frequency plots of the crRNAs-targeted sites for Trp53, Apc, Pten and Rb1 demonstrating the type of gene modification generated in SCC samples. b, Nextera sequencing quantifies percent gene modification for the four targeted genes, Trp53, Apc, Pten and Rb1, in SCC samples. Two-way ANOVA with Šídák’s multiple comparisons test was used to assess significance. Data are shown as mean ± s.e.m. P values are labelled. For all genes, N = 5 biological replicates. c, Schematic demonstrating the process of isolating single clone from primary tumour SCC/LUAD samples and surveying the efficiency of single-cell-level quadruple knockout of the four TSGs. Created in BioRender. Zhou, L. (2025) https://BioRender.com/n05e955. d, Quantification of the number of single clones with simultaneous four TSGs edited. e, Circos plot created with RCircos with variant information regarding crCD80 (blue), crH2Ab (green) and CD80-H2Ab1-DKO (red) genome. The outer section composed of three scatter plots (one per group) with wheat colour background, representing the average coverage at each 1 Mb window (blue dot, crCD80; green dot, crH2Ab1; red dot, CD80-H2Ab1-DKO). The inner section is composed of three histogram plots (one per group), representing the variants count by windows of 1 Mb (blue track, crCD80; green track, crH2Ab1; red track, CD80-H2Ab1-DKO). f, Heatmap of percent modification of each predicted off-target site for H2Ab1 crRNA. X-axis is composed of vector, crCD80, crH2Ab1 and CD80-H2Ab1-DKO, each with three biological replicates. Y-axis is composed of the surveyed genomic regions, including one on-target site and ten predicted off-target sites for crH2Ab1. g, Bar plot summarizing the percent modification in predicted off-target sites of H2Ab1 crRNA for vector, crCD80, crH2Ab1 and CD80-H2Ab1-DKO. Data are shown as mean ± s.e.m. N = 3 biological replicates for all groups. h, Heatmap of percent modification of each predicted off-target site for CD80 crRNA. X-axis is composed of vector, crCD80, crH2Ab1 and CD80-H2Ab1-DKO, each with three biological replicates. Y-axis is composed of the surveyed genomic regions, including one on-target site and nine predicted off-target sites for crCD80 (crCD80_6 not included owing to primer synthesis failure). i, Bar plot summarizing the percent modification in predicted off-target sites of CD80 crRNA for vector, crCD80, crH2Ab1 and CD80-H2Ab1-DKO. Data are shown as mean ± s.e.m. N = 3 biological replicates for all groups. Source data

Fig. 5. Simultaneous DAKO with LSL-enAsCas12a-HF1;dCas9-SPH double transgenic mice.

a, Schematic showing the breeding strategy of LSL-enAsCas12a-HF1;dCas9-SPH mice. Created in BioRender. Chen, S. (2025) https://BioRender.com/h16k998. b, Schematic demonstrating the mechanism of the simultaneous DAKO system. A dCas9-sgRNA was concatenated with Cas12a DR and crRNA. Upon delivery into the same cell, the guide chimera was then processed and cleaved into mature dCas9-sgRNA and Cas12a-crRNA by Cas12a protein. The mature dCas9-sgRNA and Cas12a-crRNA then assembled with corresponding protein to mediate gene activation and gene knockout, respectively. RNP, ribonucleoprotein. Created in BioRender. Chen, S. (2025) https://BioRender.com/t91r170. c, Schematic illustrating the retroviral construct and the workflow of ex vivo of DAKO validation in primary BMDCs. Human U6 promoter (hU6) drives the expression of a Cas9 activating guide for Itgb4 (Itgb4-sgRNA) concatenated with a Cas12a DR and crRNA targeting CD24 (crCD24). A Cre expression cassette was used to induce enAsCas12a-HF1 expression. BM, bone marrow. Created in BioRender. Chen, S. (2025) https://BioRender.com/d07o322. d, Flow cytometry analysis on LSL-enAsCas12a BMDC Itgb4-CD24 DAKO experiment. Representative contour plots of mScarlet⁺ BMDCs showing the Itgb4 expression in relation to the CD24 expression. Anti-ITGB4-APC and anti-CD24-BV421 were used in the flow staining. DAKO-crRNA containing both crItgb4 and crCD24 was compared with single crRNA (Itgb4-sgRNA and crCD24) and vector control. e, Quantification of CD24⁻, ITGB4⁺ and CD24⁻;Itgb4⁺ percentage with respect to mScarlet⁺ BMDCs. One-way ANOVA with Tukey’s multiple comparisons test was used to assess significance. For bar plot, data are shown as mean ± s.e.m. Exact P values are labelled. For all groups, N = 3 biological replicates. Source data