Four questions to Luigi Naldini (San Raffaele Telethon Institute for Gene Therapy, Milan) about the Nature Biotechnology study that revealed limitations and risks of gene and prime editing.

No platform is flawless, but is it still possible to draw up a hierarchy of reliability by saying, for example, that base and prime editing (the pencil and spell checker tool in the illustration above) generate fewer unwanted mutations than Cas9 (the scissors)?

Yes of course. DNA double-strand breaks (DSBs) are the main cause of unwanted genetic alterations that can result from editing (even very large deletions up to the loss of a chromosome arm, chromosome translocations, complex rearrangements). So the reduction in the occurrence of DSBs, although not complete, achieved by base and prime editing still improves their biosafety compared with the use of CRISPR scissors.

CRISPR-Cas9 is still the most widely used CRISPR tool because it was invented first and is very efficient. We know that in embryos it can cause worrisome macro-mutations but in adult cells it works well, so much so that a treatment for sickle cell anemia is about to receive the green light for commercialization. Does the reliability of platforms also depend on the conditions under which they operate?

We know that even just reducing the exposure time of cells to editors can increase their accuracy, reducing unwanted effects and maintaining adequate efficiency. So by refining the procedure (e.g., using the pre-assembled protein with its guide instead of delivering two separate RNAs for the editor and guide) there is room to improve its reliability. Then of course the response may depend also on the cell type and even its conditions. For example, cutting DNA in a cell that is about to divide is more dangerous than doing it in a quiescent cell. We have seen that both the efficiency of base editing and subsequent adverse events are affected by the readiness with which a cell detects a mis-match between two bases – just the one introduced by the base editor – and takes action to fix it. All of these conditions, and likely others yet to be discovered, will allow the accuracy and safety of genetic intervention to be optimized.

At SR-Tiget you have discovered the breaks created by the more advanced CRISPR platforms, how are you working now to put the patches in place?

First, by optimizing the editor’s delivery conditions based on the cell type being treated and verifying that the undesirable events we have documented are mitigated. Then, having identified the “Achilles’ heel” of these amazing molecular machines, we can direct their evolution in vitro toward safer variants.

On the one hand, one must proceed with caution; on the other hand, one cannot set unattainable safety standards. How do you find the balance point?

The history of medicine has taught us to use the risk-benefit ratio in the design of each new trial, aiming to ensure as much safety and benefit to the patient as possible, but also the possibility of gaining knowledge that will illuminate the subsequent drug development pathway. The more we know about the mechanism of action and risks associated with a new drug, the better we will be able to evaluate its justified application to a certain disease, so we will consider the severity of the disease, the urgency of treatment, and the availability of other effective or palliative treatments. Today, for some serious genetic disorders such as blood diseases, we have the luxury of choosing among several types of advanced therapies, such as replacing a functional version of the defective gene with the help of a viral vector or directly editing its sequence to restore its function or, again, awakening a vicarious gene, such as a fetal copy, that can compensate for the defect. This “choice” carries a great deal of responsibility for the research clinician, who must also weigh the greater knowledge gained over time with the more established therapies against the broader degree of “unknown” associated with newer, innovative ones.