Unless you were on Mars or in the middle of a Netflix binge, you couldn’t have missed the paper that has been hitting the headlines all over the world this week – the first example of the genetic modification of human embryos. The study led by Junjiu Huang (Sun Yat-sen University, Guangzhou) used CRISPR-Cas9 gene editing technology to attempt the repair of the HBB gene, which causes β-thalasseamia, in tri-pronuclear human zygotes (3PN). These abnormal embryos are deemed non-viable, although they can be put to use to study some aspects of human zygotic development and so were used in this “proof-of-principle” assessment. While the ethical questions associated with this study have deeply divided the scientific world, let’s take a chance to look at the science behind the headline.
The group attempted gene repair in 86 3PN embryos:
- At 48 hours, 71 (82.6%) 8-cell stage embryos remained viable.
- Of the 54 embryos that could be analyzed, only 28 were cleaved by Cas9 (~52% efficiency).
- The error prone non-homologous end joining (NHEJ) pathway directed ssDNA-mediated gene editing in 4 embryos (14.3%).
- The high fidelity non-crossover homology directed repair (HDR) pathway using the endogenous HBD gene as a repair template and not the ssDNA HBB donor template edited 7 embryos (25%).
- Both these findings contribute to the introduction of mutations and mosaicism of HBB allele presence in the 3PN embryos.
- Whole exome sequencing of HBB cleaved embryos demonstrated multiple off-target mutations.
- Whole genome sequencing would likely show further mutations.
The low efficiency repair and significant off-target effects observed in this study suggest that this current CRISPR-Cas9-mediated strategy cannot be safely used to mediate effective gene repair in viable human embryos. The same strategy generally functions at high efficiency in human cells and as a control in this study, the authors found that the same strategy utilized in human embryonic kidney (293T) cells gave efficient cleavage activities. They found no adverse effects in the top 7 predicted off-target sites in the genome, or at the HBD locus, alongside a ~50% repair success rate while using a competing ssDNA oligo donor template. This suggests that either the presence of 3 pro-nuclei or the developmental stage of the embryos may disrupt the CRISPR-Cas9-mediated repair process. The combination of these findings with the known propensity of CRISPR-Cas9 to have off-target effects (ChIP Shows Cas9 Can Miss the Mark, The Latest and Greatest: A Genome-Editing Cautionary Tale, and Keeping Tabs on Off-Target Effects of Genome Editing) suggests that gene editing in human embryos may require a change of tactic to become a safe and effective strategy.
See the paper that is making all the headlines in Protein and Cell, 2015.