Homozygous editing of multiple genes for accelerated generation of xenotransplantation pigs

Abstract

Although CRISPR-Cas-based genome editing has made significant strides over the past decade, achieving simultaneous homozygous gene editing of multiple targets in primary cells remains a significant challenge. In this study, we optimized a coselection strategy to enhance homozygous gene editing rates in the genomes of primary porcine fetal fibroblasts (PFFs). The strategy utilizes the expression of a surrogate reporter (eGFP) to select for cells with the highest reporter expression, thereby improving editing efficiency. For simultaneous multigene editing, we targeted the most challenging site for selection, whereas other target sites did not require selection. Using this approach, we successfully obtained single-cell PFF clones (three of 10) with seven or more homozygously edited genes, including GGTA1, CMAH, B4GALNT2, CD46, CD47, THBD, and GHR Importantly, cells edited using this strategy can be efficiently used for somatic cell nuclear transfer (SCNT) to generate healthy xenotransplantation pigs in <5 months, a process that previously required years of breeding or multiple rounds of SCNT.

Figure 1.

Detection of targeting events using surrogate reporters. (A) The principle of the surrogate reporter system. The reporter gene is initially frameshifted owing to the insertion of a CRISPR target site adjacent to the “ATG” start codon. CRISPR-Cas9 mediated editing results in insertions and deletions (indels) that restore the correct reading frame of the reporter gene. (B) Workflow for assessing the correlation between genomic and plasmid DNA targeting at the single-cell level. Colonies containing an integrated surrogate reporter (eRFP) were electro-transfected with Cas9, sgRNA, and a surrogate reporter (eGFP) plasmid. Both genomic (eRFP) and plasmid (eGFP) surrogate reporters incorporate the same target site (TS1). (ITR) Inverted terminal repeat sequence. (C) Flow cytometry analysis distinguishing colonies edited in genomic DNA (PE positive) and plasmid DNA (FITC positive). The selection efficiency is calculated by the ratio of GFP and RFP double-positive cells to GFP-positive cells. (D) Workflow for assessing the correlation between endogenous and surrogate DNA targeting at the single-cell clone level. Cells were electro-transfected with Cas9, sgRNAs, a surrogate reporter (Puro), and PBase. Following puromycin selection, cell colonies were picked for genotyping. (E) Primers (F1/R1) amplify the surrogate target site (Puro) integrated into the genome, and primers (F2/R2) amplify the endogenous locus. The right panel illustrates the ratio of edited clones.

Figure 2.

Efficient enrichment of genome-targeted cells using a surrogate reporter. (A) Enrichment of genome-targeted cells using a surrogate reporter (Puro). HeLa cells were cotransfected with Cas9, sgRNAs, a surrogate reporter (Puro), GFP, and PBase plasmids. Transfection-positive cells (GFP⁺) were divided into two groups: One group was treated with a high dose (2 µg/mL) of puromycin for 5 days, and the other group remained untreated. Targeting efficiency of these two group cells was determined by ICE analysis from Sanger sequencing. Cell colonies were picked for further genotyping. (**) P ≤ 0.01, (***) P ≤ 0.001 (Student's two-tailed t-test). (B) Enrichment of CMAH- and GGTA1-targeted PFFs by surrogate reporter (Neo) and surrogate reporter (GFP), respectively. (C) Enrichment of Tet2-targeted mESCs by surrogate reporter (Neo) and of et1-, Tet2-, and Tet3-targeted mESCs by surrogate reporter (eGFP), respectively. (D) The GFP expression in genomic targeted and nontargeted mESCs clones selected by the surrogate reporter (eGFP) plasmid. Scale bars, 1000 μm.

Figure 3.

Selection of homozygously edited cells based on highest surrogate reporter expression. (A) Schematic of selection strategy using version 2.1-eGFP (V2.1G), a nonintegrated surrogate reporter system. PFFs were cotransfected with V2.1G, Cas9, and sgRNA plasmid, whereas the control group was only transfected with the V2.1G plasmid. FACS analysis revealed varying intensities of FITC expression among the cells. (H) The highest FITC intensity population, (M) the middle FITC intensity population, (L) the lower FITC intensity population, and (N) the negative FITC intensity population. (B) Correlation between genome editing efficiency and FITC intensity of endogenous targeting in PFFs. (C) Representative Sanger sequences of CMAH target site in wild-type cells, homozygous cell clones, and cell pools selected by V2.1G. (D) FACS analysis of GGTA1 targeting in PFF pools detection by IB4 lectin staining. (E) Western blot analysis of USE1-targeted Vero E6 cell pools with middle and high FITC intensity, respectively.

Figure 4.

Simultaneous targeting of seven genes using V2.1G. (A) Schematic for targeting seven endogenous genes in PFFs using V2.1G. A surrogate reporter (eGFP) plasmid is used for selecting B4GALNT2-targeting cells. After 24 h of transfection, cells were sorted into three groups: V2.1G (FITC-highest), Enrich (PE-highest), and Control (PE-positive), for genotyping and colony formation. (B) Genotyping of the clones for seven genes at 16 alleles. Each box represents one allele, filled boxes represent a targeted allele, and empty boxes represent a nontargeted allele.

Figure 5.

Enhanced knockout and knock-in efficiency through coselection strategy. (A) Generation of PFF clones with GGTA1 and CMAH knockouts, and hTHBD and hCD55 knock-ins via Cas9-mediated NHEJ. The target sites, located in the first exon of GGTA1 and CMAH, are indicated by arrows. Puro and Neo serve as surrogate reporters for GGTA1 and CMAH targeting, respectively. The EF1a-hTHBD-P2A-hCD55-pA sequence is used for insertion. Clones are categorized as follows: (Homo + Hete) one gene homozygously edited, one heterozygously edited in GGTA1 and CMAH, (Homo) both genes homozygously edited, and (Homo + KI) both genes homozygously edited with hTHBD and hCD55 knock-ins. (B) Generation of PFF clones with triple knockout of GGTA1, B4GALNT2, and CMAH and with knock-in of six genes—hHOMX1, hCD55, hCD47, hCD46, hTHBD, and hEPCR—via Cas9-mediated HDR. V2.1G is used for B4GALNT2 targeting. The CMV-hHOMX1-T2A-hCD55-EF1a-hCD47-P2A-hCD46-pA cassette is inserted into the GGTA1 locus, and the ICAM2-hTHBD-P2A-hEPCR-pA cassette is inserted into the CMAH locus, using ∼1 kb of homologous arms. (Right) Junction PCR for V2.1G-selected (V; V2.1G) and control (C; control) cell pools. Control pools consist of transfected positive cells. Primers detect left and right junctions in CMAH and GGTA1 loci, with “GAPDH” as a normalization control. (C) Analysis of triple gene knockout in “3KO + 6KI” cell pools. (D) Ratio of triple knockout and six genes knock-in PFF clones in V2.1G-selected and control groups. BM-PFFs♂ and BM-PFFs♀ represent PFFs from male and female Bama miniature pigs, respectively. WZS-PFFs♂ indicates PFFs from male Wuzhishan pigs.

Figure 6.

One-step generation of multiplex genome editing pigs for xenotransplantation using coselection strategy. (A) Schematic for generating five genes editing xenotransplantation pigs. (Left) The gene-targeting strategy for triple gene knockout (GGTA1, B4GALNT2, and CMAH) and dual gene knock-in (hTHBD and hCD55) in PFFs. The target sites, located in the first exon of GGTA1 and CMAH and the third exon of B4GALNT2, are indicated by arrows. The V2.1G surrogate plasmid is used to select for B4GALNT2 targeting, with selected cells serving as donors for somatic cell nuclear transfer (SCNT) to produce 5Gene xeno-pigs. (B, left) Overview of the DNA sequences at the modification loci in the piglets shown in the right panel. (C) FACS analysis validating the three-gene knockout (3KO) in peripheral blood mononuclear cells (PBMCs). (D) RT-qPCR validation of the two genes knock-in (2KI) in heart and kidney tissues. (E,F) Immunofluorescence (IF) validation of human thrombomodulin (hTBM) and human decay-accelerating factor (hCD55) expression in the kidney (E) and heart (F) of 5Gene pigs.