CRISPR: You Have Seen the Good, Now See the Bad

Don’t get me wrong, as a Ph.D. student in Genetics and Genomics program, my lab life without CRISPR would probably be orders of magnitude more difficult. No matter what kind of project I consider throwing myself into, I automatically have to consider using CRISPR either as a tool to derive experimental models or an experimental tool in a different sense altogether.

CRISPR is an essential tool for most molecular biologists — it offers high precision, it’s easy to use, freakishly cheap, extremely adjustable (you can target any one region in the genome, or multiple regions simultaneously), and amazingly versatile (you can use it to edit genome, to localize RNAs in the cell, label specific nucleic acids or the surrounding molecules, modulate gene expression, etcetera, etcetera).

One of the most promising and exciting applications of CRISPR is that in medicine. The high editing precision and efficiency, alongside the apparent absence of severe adverse effects in cell and animal models have resulted in high hopes that CRISPR would be extremely useful (and profitable) for treatment of, and maybe even curing, genetic diseases. Therefore, it is not surprising that the pioneers of CRISPR (as a gene editing tool) have founded companies (CRISPR Therapeutics and Editas Medicine) aiming to develop therapies using CRISPR–Cas9 gene editing technology.

These promising applications of CRISPR technologies are also why scientific publications mentioning “CRISPR” have skyrocketed over the last 10 years (Figure 1). Unfortunately, however, if it sounds too good to be true — it probably is.

In May of 2017, a bombshell paper in Nature Methods claimed CRISPR resulted in unexpected (and rather frequent, >1,000) mutations when CRISPR–Cas9 editing was attempted in vivo. However, soon after the paper was published, concerns have been raised over possibly faulty experimental design and data analysis ultimately resulting in retraction almost a year later.

Notwithstanding, just 4 months later, a new paper (this time with a sound experimental design and data analysis) was published in Nature Methods that showed there was indeed some evidence of Cas9 off-target activity (albeit not as drastic as Schaefer et al., 2017 have claimed). The same paper also made suggestions on how to efficiently reduce this activity.

Moreover, two different papers were published earlier in June demonstrating CRISPR-mediated induction of p53-mediated DNA damage response and, additionally, p53-mediated inhibition CRISPR–Cas9 activity. p53 is an important protein involved in the regulation of the cell cycle, which also makes for its very frequent involvement in cancer. These findings were important because they have essentially demonstrated that normally functioning stem cells are resistant to CRISPR-mediated gene editing. However, it has been suggested that temporary reduction of p53 may alleviate this issue.

Ultimately, I would like to share my own experience with CRISPR. It’s not as easy as it gets portrayed. Sure, it beats alternatives by multiple orders of magnitude. Nonetheless, CRISPR is not 100% efficient, and it is not 100% accurate. There is a serial reduction in the number of “successes” throughout the editing process: only a fraction of the cells successfully take in the necessary components, only a fraction of those cells result in successful cutting, and only a tiny fraction of those cells result in the desired type of editing event.

This low rate of efficiency is detrimental for some of the CRISPR’s highly anticipated medical applications. There is also the problem of delivery, with many genetic diseases being a result of the pathological state of difficult-to-reach cell populations— such as neurons in the brain.

So, you’ve seen the good — CRISPR is a truly remarkable molecular tool, it has aided the scientific community immensely. However, it’s time you also see the bad — CRISPR is not a miracle cure-all; off-target effects, difficult-delivery, imperfect editing efficiency, and on-target accuracy, in addition to interaction with cell’s essential checkpoint regulators, are all important signs that we should proceed with caution when evaluating the use of CRISPR in medical applications.

This is not to say that we should give up in pursuit of using CRISPR for treating conditions that affect the quality of life of so many people. The technology is, after all, astoundingly promising. We should, however, be cautious and not get carried away in our excitement, subsequently overlooking a possibly detrimental side-effect. We should definitely not use CRISPR to introduce unnecessary mutations into the germline, the effect of which we don’t fully understand. After all, as one of the most important pillars of bioethics teaches us — primum non nocere.