Abstract Study question Does CRISPR/Cas9 gene editing have a distinct genetic impact in human embryos compared to induced pluripotent stem cells (iPSCs)? Summary answer IPSCs display a difference in repair events, such as template usage, compared to the germline which indicates a distinct genetic impact by CRISPR/Cas9 gene editing. What is known already CRISPR/Cas9 is utilized to induce targeted DNA editing. To introduce specific changes, such as mutational correction, a DNA template is generally administered, stimulating homology-directed repair. Recent human germline editing studies aiming for mutational repair created a moderate amount of embryos which solely carried wild-type alleles. Remarkably, no template use was observed, but instead loss-of-heterozygosity (LOH) was demonstrated. This is the presence of only one (the wild-type) allele, either caused by gene conversion events using the wild type allele or chromosome loss. Up to date, no iPSCs studies have evaluated LOH events when attempting to correct a heterozygous mutation. Study design, size, duration A guide RNA targeting the mutant allele, and a repair template containing the wild-type allele and a synonymous variant to track template usage were designed. For iPSCs targeting, the components were nucleofected after which DNA was extracted from the whole well (n = 3) or from individual colonies (n = 33). For human embryo targeting, the components were injected simultaneously with sperm into donated spare oocytes (n = 32). DNA was extracted from embryos after 3-6 days of in vitro culture. Participants/materials, setting, methods Sperm and renal cells, from which the iPSCs were derived, were donated by a patient with a heterozygous base pair substitution in PLCZ1 (c.136-1G>C) causing fertilization failure. To overcome fertilization failure, assisted oocyte activation was employed during ICSI. For the embryos, next-generation sequencing (NGS), short tandem repeat (STR) analysis and a whole genome single nucleotide polymorphism (SNP) assay were performed. For the iPSCs, NGS and a targeted SNP assay were carried out. Main results and the role of chance The genetic events in embryos originating from mutant sperm (n = 32) displayed following distribution: 19% showed the untargeted mutant allele, 56% showed additional mutagenesis and 25% showed only the wild-type allele. In the latter group, the template was never utilized, pointing to LOH. STR analysis indeed revealed LOH events of different lengths in these embryos. SNP analysis of one embryo, originating from mutant sperm but displaying the wild-type allele, did not demonstrate LOH (with a detection limit of 500bp). These findings suggest the occurrence of gene conversion, which rejects the formerly stated hypothesis that the observed LOH in human embryos can mostly be attributed to chromosome loss. In the iPSCs, whole-well (n = 3) and single colony (n = 33) data demonstrated a similar trend in genetic event distribution compared to the embryos, with the main difference that in 30% of the corrected reads/colonies, template use was observed. LOH was further analysed in nineteen single colonies. In all the colonies (3/3) corrected with the template, no LOH was present. When the colonies displayed additional mutagenesis, 7% (1/13) contained LOH events. In 33% (1/3) of the colonies showing only the wild-type allele (without signs of template use), LOH was observed. Limitations, reasons for caution SNP assays have a higher resolution compared to STR analysis, and enable distinction of LOH events between ‘gene conversion’ and ‘chromosome loss’. Therefore, more embryos and iPSC colonies will be analysed with SNP assays. However, informative SNPs define the resolution of our assay (currently 500 bp from Cas9 cut site). Wider implications of the findings Our results show that template usage seems to differ between human embryos and iPSCs, but that iPSCs do not solely rely on template use as well. This demonstrates that embryo experiments to study the genetic impact of CRISPR/Cas9 cannot fully be replaced by iPSCs experiments. Trial registration number Not Applicable
Read full abstract