• The effect of CRISPR-Cas9 gene editing on chromosomal damage of human embryos
  • Fereshteh Moheb Afzali,1,*
    1. Department of Cellular and Molecular Biology, Faculty of Advanced Sciences and Technology, Tehran Medical Sciences Branch, Islamic Azad University, Tehran, Iran.


  • Introduction: CRISPR-Cas9 genome editing is a promising technique for clinical applications. In this study, computational pipelines to assess single-cell genomics and transcriptomics datasets from OCT4 (POU5F1)CRISPR-Cas9-targeted and Cas9-only were controlled human preimplantation embryos. This allowed us to evaluate on-target mutations that would be missed by more conventional genotyping techniques.
  • Methods: Ethics statement Human embryos were donated to the previous project by informed consent under the UK Human Fertilization and Embryo Authority (HFEA) License. CRISPR-Cas9 targeting of POU5F1 in vitro fertilized zygotes were microinjected with either a sgRNA-Cas9 ribonucleoprotein complex or with Cas9 protein alone and cultured for 5-6 days. The sgRNA was designed to target exon2 of the POU5F1 gene. Genomic DNA from Cas9 control and OCT4-targeted human embryos was isolated from either an individual single cell or a cluster of 2-5 cells from trophectoderm biopsies from embryos that developed to the blastocyst stage, as well as blastomeres from earlier stage embryos. Cytogenetic analysis To determine the chromosome copy number of samples, their genomic DNA was subjected to low-pass whole-genome sequencing. Libraries were prepared using the VeriSeq PGS Kit(Illumina)and sequenced with the MiSeq platform. PCR primer design and testing To independently investigate the prevalence of LOH events designed PCR primer pairs to amplify 15 fragments spanning a~20kb region containing the POU5F1 locus . also designed a control primer pair in exon 4 of the gene ARGFX, which is on a different chromosome. PCR amplification In preparation for PCR amplification and to expedite the processing of 2192 samples used the QIAgility robot for master mix preparation. The PCR reaction was run with the thermocycling settings. Targeted deep sequencing Clean PCR amplicons from the same DNA sample were barcoded and pooled to generate 137 barcoded libraries that were submitted for targeted deep sequencing by Illumina MiSeq v3(300bp paired-end reads). SNP-typing the MiSeq paired-end reads with DADA2 to remove low-quality regions trimmed. SNP calling was performed with BCFtools and call algorithms in multi-threaded mode. scRNA-seq data analysis scRNA-seq reads from G&T-seq samples were aligned to the human reference genome GRCh38 using TopHat2. selected samples for our scRNA-seq analyses are based on the quality of the eSNP-karyotyping profiles.
  • Results: Segmental losses and gains at a CRISPR-Cas9 on-target site identified by cytogenetics analysis The number of segmental and whole-chromosome abnormalities observed in the CRISPR-Cas9 targeted human cells was significantly higher than in the control samples(P = 0.0405). Moreover, this significant difference can be attributed to the observed segmental abnormalities on 6p, because excluding them from the comparison results in a negligible difference in whole-chromosome abnormalities between targeted and control samples(P = 0.2419). Loss-of-heterozygosity identified by targeted deep sequencing data led to the identification of four different patterns: samples without LOH, that we're able to call heterozygous SNPs in multiple amplified fragments. Cases with putative LOH at the locus have heterozygous SNPs in the amplicons covering exons 1 and 5 of the POU5F1 gene(fragments E1-2, G1, and E4)and homozygous SNPs that could represent CRISPR-Cas9-induced deletions in the order of~4kb. Bookended samples have two heterozygous SNPs flanking the cut site but in fragments outside the POU5F1 locus and could represent deletions of lengths between~7kb and~12kb. Finally, in open-ended samples, it was not possible to find heterozygous SNPs in any of the amplified fragments or there was one or a few heterozygous SNPs on only one side of the region of interest that could represent large deletions of at least~20kb in length. No evidence of on-target complexity using digital karyotype and LOH analysis of the single-cell transcriptome data.
  • Conclusion: in contrast to Cas9 control embryos, a significantly higher frequency of CRISPR-Cas9 targeted embryos with a segmental gain or loss that was directly adjacent to the POU5F1 on-target site. LOH events may be overestimated.55.6% of CRISPR-Cas9 targeted cells did not exhibit any obvious segmental or whole chromosome 6 abnormalities and their genotype and phenotype, concerning the OCT4 function, are interpretable. Given the likelihood of mosaicism, it is unclear whether the segmental abnormalities we observed in any one cell analyzed from each embryo are representative of the entire CRISPR-Cas9 targeted embryo or a subset of cells within the embryo. overall, this points to the need to develop a robust technique to distinguish cells and embryos affected by CRISPR-Cas9 unintended damage from correctly edited embryos.
  • Keywords: CRISPR-Cas9, genome editing, chromosome, loss- of heterozygosity, human embryos