Evolutionary engineer Kevin Esvelt, at the Massachusetts Institute of Technology in Cambridge, works with gene drives, engineered bits of DNA that can cause a mutation to become heritable all the time. He calls for researchers to create and use safe lab procedures while working with this powerful but potentially risky technology.

What is a gene drive?

In nature, a gene drive occurs when a DNA sequence spreads through a population by breaking the conventional rules of inheritance. For example, if an organism has a single copy of a fluorescent marker gene and its mate has none, normally only half their offspring will fluoresce. When a gene-drive system is in play, almost all of them will glow.

How can scientists use this capability?

Gene drives allow us to drive altered traits through wild populations over generations. For instance, we could alter the DNA of wild mosquitoes to stop them from carrying disease. We could restore damaged ecosystems and save endangered wildlife by genetically removing invasive species.

How did your insights help to propel this field?

Even ten years ago, heritable genome editing was a possibility, but no one had found a molecular tool that would enable it to be done efficiently. In 2013, laboratories began using CRISPR to precisely edit the genomes of many species. I realized then that this tool could be used to build stable gene drives in many complex organisms. It could also be used to build reverse drives, which are like molecular erasers for overwriting previous edits.

Why did you explain how gene drives would work before you published results showing that they could work in any organism?

Most advances don't give individual scientists the power to affect entire ecosystems. By detailing what was possible, how it could be achieved and what safeguards were needed to prevent any accidental release of altered organisms from the lab, we hoped to set an example of how future work in gene drives should proceed.

Credit: Erika DeBenedictis/MIT Media Lab

Why was this important?

A single escaped organism that found a mate could eventually alter most of the local population and, very possibly, every population of that species worldwide. The ecological risk might be low, but the damage to public trust in biotechnology could imperil the future of the field.

Did you want researchers to agree on some guidelines first?

My immediate priority was to prevent the accidental release of any gene-drive organisms into the wild. I wrote to the few researchers working on gene drives to explain my concerns about ethics and safety.

What happened?

Last year, we released results showing that gene drives work in yeast. Then another group — who were working with fruit flies — independently created a functional gene-drive system. They were careful to keep the flies contained, but unlike our paper, their manuscript, which was meant to be published as a how-to for other labs, made no mention of safeguards or the risk to wild populations. To their credit, they agreed to include those details.

Did your efforts help to usher in regulation?

The fruit-fly case triggered responses from many scientists. For months, we struggled to agree on which safeguards should be used in the lab. We eventually published our recommendations in July 2015, and this year the US National Academy of Sciences released a report setting out how to conduct gene-drive research responsibly.

Should gene-drive information be classified?

Classifying such information would hinder beneficial applications and threaten biosecurity. We must know which species to monitor. Open science is the best defence and the best way to earn public support.

This interview has been edited for length and clarity.