Promoting Cas9 degradation reduces mosaic mutations in non-human primate embryos

Abstract

CRISPR-Cas9 is a powerful new tool for genome editing, but this technique creates mosaic mutations that affect the efficiency and precision of its ability to edit the genome. Reducing mosaic mutations is particularly important for gene therapy and precision genome editing. Although the mechanisms underlying the CRSIPR/Cas9-mediated mosaic mutations remain elusive, the prolonged expression and activity of Cas9 in embryos could contribute to mosaicism in DNA mutations. Here we report that tagging Cas9 with ubiquitin-proteasomal degradation signals can facilitate the degradation of Cas9 in non-human primate embryos. Using embryo-splitting approach, we found that shortening the half-life of Cas9 in fertilized zygotes reduces mosaic mutations and increases its ability to modify genomes in non-human primate embryos. Also, injection of modified Cas9 in one-cell embryos leads to live monkeys with the targeted gene modifications. Our findings suggest that modifying Cas9 activity can be an effective strategy to enhance precision genome editing.

Figure 1. Targeting Ubiquitin to Cas9 to Increase Its Degradation.

(A) Schematic of the Ubiquitin-Cas9 (Ubi-Cas9) vector for the expression of Cas9 with short half-life. Red box indicates the inserted ubiquitin-targeting signal that is tagged to the N-terminus of Cas9. Green box indicates T7 promoter. The C-terminus of Cas9 carries the nuclear localization signal (NLS). (B) Western blot analysis for comparing the half-life of WT-Cas9 and Ubi-Cas9 in transfected HEK293 cells that were treated with 50 μM cycloheximide for different times (0, 4, 8, 12, 24 h) to inhibit protein synthesis. The relative protein levels of WT-Cas9 and Ubi-Cas9 were assessed by densitometric analysis of their bands on western blots, and the value at 0 h was considered 100% (mean ± SEM; n = 3, *p < 0.05; **p < 0.01). Images of full-length gels are seen in supplemental information. (C) Immunocytochemical assays showing the rapid degradation of Ubi-Cas9 in monkey embryos after injection of Ubi-Cas9 mRNA for 24 h or 40 h. WT-Cas9 mRNA injection served as controls. Embryos were stained with anti-Cas9 (green) and DAPI (blue) for nuclear staining. Note that more Ubi-Cas9 is located in the nucleus than the cytoplasm. Scale bar: 20 μm.

Figure 2. Comparison of Pink1 DNA Targeting by Injection of Cas9 mRNA and Cas9 Protein.

Identification of Pink1 DNA mutations by T7E1 assay in monkey embryos injected with WT-Cas9 or Ubi-Cas9 mRNA (A) or purified WT-Cas9 protein (B). Representative DNA gel images showing PCR results of single cells from different embryos. Arrow indicates DNA mutation in a single cell from 8-cell embryos. PC is a positive control from embryos injected with Cas9 mRNA and gRNA for Pink1 targeting. M: Molecular markers. (C) Comparison of targeting rates of injected Cas9 protein, WT-Cas9 mRNA, and Ubi-Cas9 mRNA in monkey embryos. The numbers of embryos with targeted gene mutations or embryos containing mutations in all cells are presented. Embryos at the 4- and 8-cell stages were used, and data were obtained from six experiments.

Figure 3. Use of Embryo Splitting to Assess Ubi-Cas9 Targeting Efficiency.

(A) Embryo splitting at the 4-cell stage to separate single cells (I, II, III, and IV), which were placed into individual empty zonae pellucidae to further develop to divided embryos that consist of 4 cells (1, 2, 3, and 4). Single cells (1, 2, 3, and 4) were then isolated and used for PCR analysis. (B) Photographs of development of divided monkey embryos: a single cell isolated from 4-cell embryos was transferred into an empty zona pellucida, and then developed to 2- or 4-cell or hatched blastocyst. (C) Hpy188I digestion of the targeted Pink1 gene showing that Ubi-Cas9 mRNA injection yielded a high rate of homogenous mutations in monkey embryos. PCR products of single cells (1, 2, 3, 4) from divided embryos were analyzed. For Hpy188I digestion, red arrows indicate biallelic mutations, whereas green arrows indicate a single allele mutation. Images of full-length gels are seen in supplemental information. (D) In cultured blastomeres that were isolated from embryos injected with Ubi-Cas9 or WT-Cas9, Ubi-Cas9 yielded biallelic mutations in 53% of blastomeres whereas WT-Cas9 generated bialleic mutations in 28% of balstomeres.

Figure 4. Ubi-Cas9 Targeting of Pink1 Gene.

(A) A diagram showing single cell PCR using intact or divided monkey embryos. (B,C) PCR diagnosis of Pink1 exon 2 targeting via Ubi-Cas9 in single cells from intact embryos (B) or divided embryos (C). Gene mutations were diagnosed via digestion of PCR products by Hpy188I. Green arrows indicate one allele mutation, and red arrows indicate biallelic mutations. (D) Quantitative assessment of Ubi-Ca9-mediated homogenous mutations in single cells from intact embryos at the 4–8-cell stage or from divided embryos derived from 4-cell intact embryos. In this graph, 31 intact embryos were used to isolate single cells, and 13 4-cell-stage embryos were divided to single cells for further development to 37 single blastomere embryos; data are presented as mean ± SEM. **p < 0.01. (E) DNA sequencing verified the homogenous mutations in single cells from divided embryos injected with Ubi-Cas9 and gRNA for Pink1 exon 2.

Figure 5. Mutations of the ASPM Gene in Monkey Tissues.

(A) T7E1 digestion of the targeted ASPM gene in embryos injected with WT-Cas9 or Ubi-Cas9. Green arrows indicate the mutations in single cells isolated from 2-cell or 4-cell embryos. (B) ASPM immunofluorescent staining of 2-cell-stage monkey embryos showing that injection of Ubi-Cas9 mRNA and gRNAs for targeting ASPM can deplete the spindle localization (arrows) of ASPM. Scale bar: 20 μm. (C) Sequence analysis of tissue DNAs from a live monkey indicating mutations in the ASPM gene that was targeted by two gRNAs for exon 3 and exon 10. The numbers of cloned DNA for sequencing and mutated DNA clones are indicated in parentheses (mutated clones/total clones). Targeted sequence regions are indicated in red.