Frequent Gene Conversion in Human Embryos Induced by Double Strand Breaks

Abstract

Applications of genome editing ultimately depend on DNA repair triggered by targeted doublestrand breaks (DSBs). However, repair mechanisms in human cells remain poorly understood and vary across different cell types. Here we report that DSBs selectively induced on a mutant allele in heterozygous human embryos are repaired by gene conversion using an intact wildtype homolog as a template in up to 40% of targeted embryos. We also show that targeting of homozygous loci facilitates an interplay of non-homologous end joining (NHEJ) and gene conversion and results in embryos which carry identical indel mutations on both loci. Additionally, conversion tracks may expand bidirectionally well beyond the target region leading to an extensive loss of heterozygosity (LOH). Our study demonstrates that gene conversion and NHEJ are two major DNA DSB repair mechanisms in preimplantation human embryos. While gene conversion could be applicable for gene correction, extensive LOH presents a serious safety concern.

Frequent loss-of-heterozygosity in CRISPR-Cas9-edited early human embryos

“…Based on our data, the possibility of gene editing via IH-HR cannot be definitely excluded. A pre-print by Liang et al (36) suggests that IH-HR could be one of the major DNA double-strand break repair pathways in human embryos. Following a similar approach to their previous study (8), the authors used CRISPR-Cas9-mediated genome editing to target a paternal mutation and were able to amplify an ~8kb genomic DNA fragment which, together with G-banding and FISH of ESCs derived from targeted embryos, suggests that repair from the maternal chromosome by IH-HR results in a stretch of LOH.…”

Severe perinatal hydrops fetalis in genome edited pigs with a biallelic five base pair deletion of the Marfan syndrome gene, FBN1

“…We present a case of an unanticipated phenotype associated with CRISPR-Cas9 gene editing. We have previously reported the reinsertion of a deleted fragment close to the CRISPR target site in a mouse line (Summers et al, 2019b) and others have presented similar outcomes in edited human embryos, including loss of whole chromosomes (Alanis-Lobato et al, 2020; Liang et al, 2020; Zuccaro et al, 2020). It is not clear whether the phenotype in these piglets was due to an on-target event similar to these or an off-target event (although this was predicted to have a low probability by the guide design software), perhaps affecting the nearby GLDN gene, as we did not subject the DNA to whole genome sequencing.…”

“…Hydrops was recently seen in an infant with biallelic mutations in the gliomedin gene, GLDN , located 3 megabases downstream of FBN1 (Australian Genomics Health Alliance Acute Care Flagship, 2020; see Table 1 ). There have been a number of reports of DNA changes in the vicinity of a target sequence associated with the use of CRISPR-Cas9 gene editing (Summers et al, 2019; Alanis-Lobato et al, 2020; Liang et al, 2020; Zuccaro et al, 2020). It is possible that GLDN was mutated by the CRISPR-Cas9 process and is responsible for the hydrops fetalis phenotype in our piglets.…”

Human Germline and Heritable Genome Editing: The Global Policy Landscape

“…A clear and accurate understanding of the global policy landscape for heritable human genome editing is especially important in light of several recent developments, including the He Jiankui debacle 33 ; reports of other researchers stating that they intend to conduct clinical experiments aimed at producing genomeedited children 34 ; ongoing debate about the need for a global moratorium 32,[35][36][37] ; growing public and policymaker interest in, and concerns about, such experiments 38,39 ; recently published policy considerations on human genome editing 36,[40][41][42] ; advances in genome editing in human embryos, which confirm significant safety concerns [43][44][45][46] ; the recent report Heritable Human 47 To promote clarity and accuracy, this survey provides an overview of the existing global policy landscape for both heritable genome editing and germline genome editing-two related research practices with differing aims and consequences. Heritable genome editing involves the transfer of genetically modified embryos to a uterus to initiate a pregnancy that would result in the birth of a child with a modified genome.…”