Keep CRISPR Safe

The possibility of rewriting the genome of an organism, or even of an entire species, has long been the stuff of science fiction. But with the development of CRISPR (which stands for “clustered regularly interspaced short palindromic repeats”), a method for editing DNA far more precisely and efficiently than was possible with older technologies, fiction has edged closer to reality. CRISPR exploits an ancient system that allows bacteria to acquire immunity from viruses. It uses an enzyme called Cas9 to cut strands of DNA at precisely targeted locations, allowing researchers to insert new genetic material into the gap. 

CRISPR promises to revolutionize gene editing, which comprises two distinct but related fields. The first involves a technique to modify inherited genes in nonhuman organisms in order to spread a trait throughout a population, using a process known as a gene drive. The other involves editing the human genome, either in normal body cells (known as somatic cells) or in the germline, the cells that pass genes down to offspring.

The advances made possible by CRISPR could bring vast benefits to society, but the technology also poses risks. An out-of-control gene drive could drastically alter or even threaten a species. And editing the human genome raises risks both for individuals and for society as a whole. To head off those dangers, governments and scientific institutions will have to respond by establishing standards that both enable promising research to go forward and reassure the public that the work is being conducted responsibly. Yet especially when the science is at such an early stage, there is a risk that governments will do too much rather than too little. To avoid that problem, the global scientific and biological ethics communities must take the lead, designing standards and procedures that reduce the dangers of these powerful new technologies without forgoing the benefits.

PLAYING GOD

The goal of a gene drive is to spread or suppress certain genes in a wild population of organisms. It works by exploiting a quirk of nature. In sexually reproducing species, most genes have a 50 percent chance of being passed from parent to child, as offspring receive half their genes from each parent. As a result, genetic mutations normally spread only if they make an organism more likely to survive or breed. But some genes have evolved mechanisms that give them better than 50 percent odds of being passed on. That allows changes in those genes to proliferate quickly even if they have no effect on evolutionary fitness. Scientists can exploit this tendency by using CRISPR to insert genetic material into the “selfish” part of an organism’s genome, ensuring that the new trait will be passed on to most offspring, eventually spreading through large populations.

The risks of CRISPR are dramatic, but they are also far off.

This process could be exploited both to improve public health and to promote economic development. Scientists could use gene drives to break disease transmission chains, eliminating the need to use costly and harmful insecticides. For example, researchers are looking at using the technology to inhibit the transmission of Lyme bacteria from mice to ticks, a move that could wipe out Lyme disease among humans, since humans can catch the disease only from tick bites. In agriculture, gene drives could immunize plants against many kinds of pathogens and curb or eliminate populations of invasive animals, such as mice, that destroy crops. Researchers are working to develop a house mouse that can give birth only to male offspring. If the altered mice were released into the wild, female mice would gradually disappear and the species would die off within the area in which the altered mice were introduced.

Gene drives could also reverse some worrying environmental trends. Many amphibian species, such as frogs, toads, and newts, have suffered catastrophic declines over the last few decades. If scientists introduced engineered genes that rendered amphibians immune to common pathogens, many species could recover.

But alongside these benefits come serious risks. A gene drive gone wrong could leave a species extinct or introduce dramatic and unintended effects. Most gene drives would likely be limited to a single place, such as an island or an isolated and containable area of land, at least at first. If a genetically modified animal or plant escaped, however, then the gene drive could spread uncontrollably. And if a modified organism mated with a member of another species, it could transmit the changes to new populations. Entire species could be wiped out and ecosystems upended.

These risks are dramatic, but they are also far off. In part because the research is at such an early stage, scientists disagree about how much and what kind of regulation and guidance will be required. So governments should follow the principle of regulatory parsimony, which dictates that they should impose only those restrictions necessary to maintain ethical standards and public safety. Doing so will maximize the scope of free scientific discovery in a way that is consistent with serving the public good. In the countries where much of the research is taking place, before imposing new restrictions, governments should rationalize the current system of regulation. In the United States, for example, the Department of Agriculture, the Food and Drug Administration, and the Environmental Protection Agency all share responsibility for different aspects of field trials and commercial products. They should harmonize their biological safety guidelines for gene-drive studies. On the international stage, the Cartagena Protocol on Biosafety, an agreement among most of the world’s countries that came into force in 2003, regulates genetically modified organisms, but the United States has not signed it. 

In lieu of formal regulations on gene drives, scientists could agree to build safety measures into gene-drive systems, such as alterations that would cancel out previous drives or gene modifications designed to grow less frequent over time, so that successive generations would express the gene less and less once the original problem has been sufficiently ameliorated. Researchers will also need to be transparent about their work and consult local communities to gain consent before introducing gene drives into the wild. 

HUMANS 2.0

Just as challenging for the global scientific community will be the issues raised by human genome editing. A wide range of diseases have been identified as potential targets for treatments that modify genes in a patient’s somatic cells, including certain cancers, cystic fibrosis, hemophilia, HIV/AIDS, Huntington’s disease, muscular dystrophy, some neurodegenerative diseases, and sickle cell anemia. Developing therapies for these conditions will not be straightforward. Preliminary laboratory research suggests that the human immune system may resist the version of the enzyme Cas9 currently used in CRISPR. If that result holds up, either Cas9 will have to be modified or replacement enzymes will have to be developed. Yet this does not appear to be a major setback, as scientists are already using other enzymes for techniques associated with CRISPR. 

Whatever form it takes, research involving human gene editing will have to meet the already rigorous regulatory standards governing medical research. In the United States, all applications for clinical experiments must be approved by the Food and Drug Administration and reviewed by the National Institutes of Health. The FDA advises researchers to follow up with participants in gene therapy trials for as long as 15 years after the end of the trial to discover and deal with any delayed ill effects. And once the FDA approves a gene therapy product for public sale, it requires companies to monitor its use, report any adverse events, and give public warnings as appropriate. This regulatory regime will be sufficient when it comes to using CRISPR to edit somatic cell genes given that the process, although different from other techniques for modifying cells, does not raise new safety or ethical issues. 

The mere fact that something can be used to do harm must not suffice to trigger regulation.

The same cannot be said for interventions that modify an individual’s germline, which carries genes that are inherited. Such interventions raise both the prospect of vast benefits and thorny questions of safety and ethics. Gene editing could, in theory, prevent the transmission of genes that increase the risks of life-threatening diseases, such as breast cancer or cystic fibrosis. Families with histories of breast cancer associated with certain mutations in the BRCA1 and BRCA2 genes (which help prevent tumors), for example, may wish to protect their descendants by editing their genes.

More speculative are germline modifications intended to make future children stronger, better looking, or smarter. The prospect of such genetic engineering raises the specter of disastrous twentieth-century experiments in eugenics, although today most of the demand would likely come from individuals rather than states. To the extent that state projects did attempt to enhance national populations, they would be ill advised and socially disruptive. What is more, because of the enormous complexity of traits such as intelligence, the results of those projects would certainly disappoint their proponents.

Reengineering the human genome raises risks not only for individual patients but also for humanity as a whole. Unlike the generations of rapidly propagating species, such as mosquitoes, human generations span many years, so any harmful change in a human germline could take decades or even centuries to become pronounced. But that does not mean that the risks should be ignored. Adjusting one part of complex human societies could well have serious consequences for public health, economic growth, and social cohesion. 

In 2017, the U.S. National Academies of Sciences, Engineering, and Medicine recommended that researchers exercise caution when it comes to efforts to prevent disease transmission through gene editing but said that such work should be allowed to go forward, albeit under “stringent oversight.” The NAS did not extend this recommendation to experiments designed to enhance future generations, which it said should not be allowed “at this time.” As the report noted, the risks of enhancement experiments are similar to those of therapeutic ones, but as long as their overall benefits are smaller, they are not worth scarce research dollars.

These sorts of guidelines will shape the work of reputable scientists, but they are not designed to stop rogue actors. At some point, governments may have to pass laws to prevent unscrupulous researchers from abusing gene editing. For now, however, the science is nowhere near advanced enough for policymakers to know what kinds of measures would work.

THE LIGHT AND THE DARK

Experiments in both human genome editing and gene drives are generally classified as “dual use”—research whose results may be used for good or evil. Since the mid-1970s, when scientists developed recombinant DNA technology, which allows DNA from different organisms to be combined to create new genetic sequences that can give organisms new traits, researchers have been concerned about the possibility of bioengineered pathogens, created either deliberately or accidentally. In 2000, researchers in Australia discovered a technique for modifying mousepox that made it more dangerous; this technique could also be applied to smallpox. In 2002, a lab in New York replicated the polio virus using publicly available DNA ordered from a biotechnology company. And in 2012, just as CRISPR was emerging, researchers in the Netherlands and the United Kingdom showed that the wild form of avian flu, which spreads from birds to mammals only through physical contact, could be genetically modified to allow it to move from birds to ferrets—and, by extension, to humans—through the air. 

The most effective standards for gene-editing research will come from the scientific community itself.

Although gene drives cannot be used to create new viruses or bacteria—neither type of pathogen reproduces sexually—they could be used to create other kinds of weapons. For example, mosquitoes could be modified to produce toxins or such that they can expand their natural habitat and so spread malaria, dengue fever, or other diseases outside tropical areas. 

Indeed, virtually all biological research could plausibly be described as dual use. It is hard to think of a major biological breakthrough that could not be exploited for harmful ends. This is one reason why all of those who value the lifesaving breakthroughs that biological research has made possible should reject the idea of regulating biological research primarily because it is dual use. Certainly, if a technology has no conceivable malign application, then regulation should be off the table. But what is equally certain, the mere fact that something can be used to do harm must not suffice to trigger regulation. 

At the moment, different countries take widely varying approaches to regulating dual-use research. The United States tends to focus only on select biological agents that threaten public health. The European Union, by contrast, takes a more precautionary approach that requires a risk assessment for any organism that could pose a threat. Regulatory parsimony and a bias toward scientific freedom favor the more focused policy, whereas greater risk aversion favors the precautionary approach. 

SELF-GOVERNMENT

For all its unprecedented power, CRISPR is of a piece with other research breakthroughs in synthetic biology. It has both enormous potential to transform societies for the better and possible malign uses. Dealing with the latter will require crafting highly specific rules so that regulators don’t end up sweeping all CRISPR research into a costly new regulatory net with little or no benefit to society. And even the best-designed regulation cannot eliminate the possibility that researchers will accidentally discover a dangerous new application of a new technology. Before regulators consider additional rules, CRISPR researchers will have to comply with existing scientific norms and regulations, perhaps the field’s biggest short-term challenge. 

The modern scientific community is both cooperative and competitive. Even so, scientific establishments have shown themselves capable of self-governance when public safety and confidence are at stake. The standards developed by recognized authorities are encouraging: for example, in the first decade of this century, in response to new laboratory practices involving the use of human stem cells in nonhuman animals, national science academies came up with a set of guidelines. The guidelines are voluntary, but they delineate in a well-informed way what is and what is not ethically acceptable, and they have been widely embraced by scientists, the editors of prominent scientific journals, and regulators.

The most effective standards for gene-editing research will come from the scientific community itself, through international summits of science academies and a continual process of intellectual exchange. Those are the forums that can respond best to often unpredictable developments in the science and react sensitively to public opinion. Prudent self-governance among scientists may not produce headlines, but it is the process most likely to enable CRISPR and the next generation of research breakthroughs to reach their full potential.