It’s Never as Simple as it Seems

As always, the annual ARVO congress had a number of hot topics. Imaging – particularly multiphoton and OCT angiography – garnered a lot of attention from delegates. One of the most oversubscribed sessions was a review of the ocular problems caused by spaceflight (1) and judging by the #ARVO2015 hashtag on Twitter, the big blue bear statue outside the Colorado Convention Center was particularly popular. As always, gene and stem cell therapies were red hot because of their huge therapeutic potential, and the results of some experiments using the whitest hot gene technology of them all, CRISPR/Cas9, was presented at the congress. Zebrafish, human cell lines, even rats in vivo, all had their genomes tweaked with this simple and readily available gene editing technique.

Of course, it’s one thing being able to achieve something pre-clinically; it’s another successfully transferring it into the clinic. And so, a fortnight before ARVO, a pretty controversial CRISPR/Cas9 paper was published (2) that highlighted just how difficult it might be to successfully deploy the technique in the clinic. A team of researchers at Sun Yat-sen University in China performed a world first: they edited the genomes of (nonviable) human embryos with CRISPR/Cas9. If we set aside the ethical debates over whether such germline-modifying procedures should be performed on embryos (notably, Nature and Science rejected the paper because of these concerns [3]), there was one thing that should dampen the gene editing technique’s luster: it didn’t work very well.

The researchers were trying to use CRISPR/Cas9 to edit the gene that encodes hemoglobin B protein (HBB). They injected 86 embryos, waited 48 hours (long enough for gene editing to occur and the embryos to grow to the eight-cell stage). Seventy-one survived; 54 were tested; only four were successfully and correctly edited – and these embryos were mosaic, meaning that only some of the eight cells were edited. Further, whole exome sequencing of these cells demonstrated a number of off-target mutations, suggesting that the CRISPR/Cas9 complex was acting on other parts of the genome too.

This might not be a killer blow for germline gene editing (ethics permitting) – the embryos weren’t “normal”; conditions can be optimized, enzymes tweaked, and there are harder-to-use but more specific alternatives to CRISPR/Cas9 like TALENs that are believed to cause fewer unintended mutations. But it is a reminder of how immature these technologies are, and how almost everything in research is more complicated than it may seem.