Thorough review argues that we are still far from using CRISPR in embryos
The list of problems includes:
- Low efficiency of in vitro fertilization procedures (successful only in 12% cases)
- High rates of aneuploidy (chromosomal abberations in 5-20% oocytes)
- Lower activity of cell cycle checkpoints (impairing standard CRISPR methods)
- Low efficacy of HDR pathway, frequent indels instead of repairs using provided template
- Widespread mosaicism (different outcomes in different cells of the same embryo)
- Large deletions near targeted region (up to 20% in one study on embryonic stem cells)
- Non-invasive sequencing is often unreliable (for instance, does not account for mosaicism)
- Insufficient assessment of off-target effects in embryos (due to limited DNA material)
Authors also describe methods which have the potential to solve the issues. Among them, there are possible optimizations of microinjection timing, addition of other molecular factors, use of modified CRISPR enzymes such as base editors, more precise detection methods.
The review proposes a framework for clinical embryo editing, which would rely on commonly agreed steps – in vitro and in vivo studies, safety and consent, as well as “broad societal consensus”.
In conclusion, they write:
Given the existing limitations in our understanding of the biology of the human embryo and how the germline genome might interact with CRISPR–Cas9-mediated DSBs, we and many others feel that the time to pursue such research clinically is not at hand.
Publication is available here: https://www.nature.com/articles/s41556-019-0424-0
Lea, R., Niakan, K. Human germline genome editing. Nat Cell Biol (2019) doi:10.1038/s41556-019-0424-0