Since the early 2000s, many advances in the life sciences, such as the artificial synthesis of the poliovirus and the gain-of-function experiments that enhanced the transmissibility of the H5N1 flu virus, have led to warnings that bioweapons development would soon be getting easier, cheaper, and faster for states and non-state actors alike. The new gene-editing technique known as Crispr has raised similar concerns because it allows researchers to edit genomes precisely, quickly, and cheaply. It has also facilitated the development of “gene drives,” which in theory allow scientists to permanently introduce a genetic alteration into an entire animal or plant population. Gene drives are being investigated as tools to eradicate infectious diseases or control pests that cause agricultural, economic, and environmental damages; yet they have also raised concerns. The absence of clear safety guidelines, coupled with ambiguous government regulations, has nurtured fears of an accidental or voluntary release of a gene drive in nature that could cause irreparable damage. On the security front, the presumed simplicity and accessibility of Crispr raise the possibility that states, terrorists, or rogue scientists might use the technology to modify genomes to develop malicious gene drives and create novel bioweapons that could spread more quickly, cheaply, and globally than traditional bioweapons agents.
Yet while it is good to be cautious about any powerful new technology, it is just as important to be realistic. From potential benefits to safety guidelines to security threats, the best approach to Crispr and gene drives will be one based on empirical findings rather than hype. Unfortunately there has been too much of the second and not enough of the first, especially when it comes to the bioweapons threat.
Understanding gene drives. A gene drive is a process by which scientists force an altered gene into an animal population to permit the inheriting of a desired trait at a higher rate and with greater certainty than through natural reproduction alone. Normally, a specific gene has a 50 percent chance of being transmitted to an offspring. Over time, however, some inherited genes may dissipate and eventually disappear from the population, particularly if they make the population less fit. With a gene drive, not only can scientists ensure that an altered gene will be transmitted with greater certainty, even if it makes the animal less fit, but also that it will spread at a faster pace across an animal population to reach an inheritance rate of nearly 100 percent.
Gene drives are not new. They may occur naturally, for example when a gene produces multiple copies of itself in a genome, or when a gene disables or destroys other genes to increase inheritance odds. Genetic alterations of domesticated plants and animals are not new, either. The novelty of recent gene drives resides in the use of the Crispr technique, which not only allows gene editing with precision, speed, and economy, but also has the potential of ensuring that alterations made in wild animals will persist in nature.
Crispr is a molecular cut-and-paste mechanism where the cut function uses small strands of RNA to guide an associated protein, such as a Cas protein, to a specific location in an organism’s DNA. The Cas protein then acts like a pair of molecular scissors to cut the DNA at this exact spot, allowing the altered gene to paste itself at that location. The technique is derived from a natural process observed in bacteria, which allows bacteria to protect themselves against viruses by cutting out sections of the viral DNA and pasting it into their own genome, thus making the bacteria immune to future viral attacks. In 2012, a team of American and European researchers found that they could use this natural process to edit genes, in theory in any organism.
With a gene drive, scientists can not only alter an animal’s gene but also insert in the animal’s genome the Crispr copy-paste system—i.e., the guide RNA and Cas protein. This copy-paste system is the gene drive, which allows the gene alteration to self-replicate in subsequent generations. For example, when an altered mosquito mates with a wild mosquito, the offspring receive an altered chromosome and a wild chromosome from each parent. The Crispr system inherited from the altered parent cuts the wild gene inherited from the wild parent and copies the altered gene into the offspring’s genome along with the gene drive. The offspring then carries two copies of the altered gene, ensuring its transmission to the next generation. When a new generation of altered mosquitoes mate with wild ones, the process repeats itself, allowing the alteration and the gene drive to spread in the whole population. The gene drive therefore appears to be a reliable mechanism for propagating altered genes, which in theory would allow gene alterations to persist in nature and permanently change the target population and possibly an entire species.
Potential benefits. Thus far, gene drives have been tested only in the laboratory, with the focus on mosquitoes that transmit infectious diseases, as well as invasive pests such as mice. The objective is to control the size of the population, suppress it entirely, or enroll it in the fight against infectious diseases. For example, a project funded by the Bill and Melinda Gates Foundation aims to use gene drives to reduce the population of malaria-transmitting mosquitoes (Anopheles gambiae) by targeting the female mosquito population that spreads the disease. Researchers are investigating two different avenues: One aims to alter the mosquito genome to decrease the number of female offspring; the second edits the genome to affect the fertility of female mosquitoes, resulting in fewer offspring. In another project, researchers at the University of California, Irvine, and Oxford University investigated the use of gene drive to produce Aedes aegypti female mosquitoes that cannot fly, potentially suppressing the local population of mosquitoes that spread dengue fever, and now the Zika virus. In yet another example, a University of California, Irvine, team used a Crispr-induced gene drive to introduce a malaria-resistant gene in the Anopheles stephensi mosquito, thus blocking the transmission of the disease, without affecting the mosquito population.
In the field of pest-control, current research involves using gene drives to limit or eradicate invasive species such as the house mouse, which can wreak agricultural havoc on land, or disturb the ecosystem, particularly on islands. Toward this end, teams at Texas A&M, North Carolina State University, and the US Department of Agriculture are working on developing a daughterless house mouse. Instead of using a Crispr system, this research exploits a natural gene drive, believed to be less potent than a Crispr-induced gene drive. The researchers hope that this feature will make it easier for regulators and the public to authorize field-testing when the time comes, particularly on islands where traditional pest-control methods have been largely ineffective. Another project proposed by MIT scientist Kevin Esvelt aims to exploit a natural gene drive to fight Lyme disease. White-footed mice have developed a natural gene drive that confers resistance to the Lyme bacteria. As ticks acquire the bacteria from the white-footed mouse, the gene drive project would act on the first ring of the tick’s food chain, thereby helping to prevent the spread of the disease to humans.
Gene drives therefore have the potential to sharply reduce the occurrence of, and possibly eradicate, several infectious diseases by disturbing their transmission chains. The rapid propagation and the long-term effects of an alteration that gene drives in theory permit suggest they are poised to be more effective that traditional disease-prevention methods, such as insecticides and rodenticides, for which animals have at times developed resistance. They could also help reduce the economic and human cost of such diseases, as well as the environmental impact of invasive species. For example, malaria kills on average 500,000 people a year, disproportionately affecting children under five years of age, and populations living in the poorest areas of Africa. The World Health Organization estimates that 390 million people are infected with dengue fever annually. And the Zika virus has been reported in 61 countries, with about 3,000 cases in the United States alone.
Gene drives also have potential to support conservation work. Gene-drive rodent control on islands can mitigate the environmental impact of invasive species, which disrupt island ecosystems by bringing in invasive plants, or eating plants and insects essential for other species’ survival. Gene drives could also possibly be used to save endangered species. On the agricultural front, gene drives could be used to control invasive species that decimate agricultural land, such as the house mouse in Australia, thereby reducing the economic cost of such annual plagues.
Safety concerns. All of this potential is not without a dark side, however, and the rapid spread of Crispr around the world has raised significant concerns about lab safety, adverse effects on target species, and potential damage to non-target species and ecosystems in general. At the moment, government regulators are struggling to keep up with the new technology, although some researchers are working to find ways to try to reduce or counteract harmful effects.
In the two years since the Crispr-induced gene-drive concept was described in a 2014 publication, gene-drive studies have been reported to cover six species. Some Crispr scientists try to monitor the spread of gene-drive research by monitoring publications and patent applications. But these two factors do not reflect accurately the spread of the technique, as some scientists may not publish until they achieve results. Scientists also disagree about the level and type of biosafety measures required by the technology. For example, should scientists use containment measures common for working with the type of organism they regularly use, or should they take additional safety measures? It doesn’t help that gene drives do not fall under a specific regulatory system. In the United States, laboratory-level gene-drive research conducted at institutions receiving federal funding must observe biosafety guidelines established by the National Institutes of Health, which identify specific safety requirements depending on the type of organism used and experiment conducted. However, these guidelines do not specifically provide guidance on containment for gene-drive research, making it difficult for biosafety committees and scientists themselves to determine whether proper safety measures have been applied.
When it comes to field trials and commercial and medical goods resulting from gene editing, these would fall under one of three different agencies depending on whether the final product is considered a food or drug (Food and Drug Administration), an insecticide (Environmental Protection Agency), or a plant pest (Department of Agriculture). In practice, however, identifying the proper regulatory agency is more complicated. For example, the Department of Agriculture has authority over genetically modified organisms, which in the 1980s and 1990s, when the regulation was developed, were modified using plant pathogens such as viruses or bacteria. But recent plant alterations made with new gene-editing techniques such as Crispr do not necessarily introduce any foreign DNA of plant pests in the altered plant, allowing newly modified crops to bypass current regulations. This is why the Department of Agriculture recently ruled that it had no authority to regulate a Crispr-edited white mushroom that resists browning. On the other hand, the Food and Drug Administration recently approved the field trial of a gene-edited mosquito produced by the company Oxitec to fight viral infections such as dengue fever and Zika in Florida, suggesting that the FDA considers such mosquitoes to be drugs, while some believe that the EPA would be a more appropriate authority to evaluate the environmental impact of edited insects.
These two cases show that Crispr-induced alterations have outpaced and continue to defy current regulations, leaving governments around the world to play catch-up. This only enhances the concerns of those who view gene editing and gene drives as a potential safety threat. In this context, fears that an altered organism might escape the laboratory to potentially eradicate a whole species, or unexpectedly jump into another population and cause unpredictable economic and environmental damage, do not seem far-fetched.
Recognizing these concerns, a group of gene-drive scientists published an article in Science last year recommending the best safety practices for gene editing in the laboratory. Scientists are also working on molecular safety measures that would control, prevent, or reverse a gene-drive in the event of a voluntary or involuntary release in nature. One such measure is a “reverse drive,” or an alteration made to an organism that would cancel out a previous drive. This measure was tested in the laboratory and showed that it could undo the genetic changes made by a previous drive in 99 percent of the cases. Another safety mechanism under development is the so-called “Daisy drive,” which places elements of a drive at different locations in the genome, so with each generation, the various elements are successively eliminated by natural selection. This mechanism would limit the spread of a drive to a specific number of generations, making the changes in a population temporary and local. Scientists have also proposed the creation of “immunization gene drives,” which would protect a species from an altered gene.
Some scientists caution, however, that safety drives will not work in 100 percent of the concerned population, and even if genetic alterations can be reversed, their effects on the ecosystem may persist. Moreover, it is unclear whether these safety mechanisms will be effective until they are tested in the field; yet testing requires oversight, which only underscores the need for a better regulatory framework.
How serious is the bioweapons threat? From a security standpoint, the rapid democratization of Crispr, the speedy results that gene drives achieve, and the perceived low cost and easy availability of laboratory equipment required for gene editing have spurred speculations that states, rogue scientists, or terrorist groups could harness the power of this new technology to spread disease-causing organisms in nature. Giving a public voice to these concerns, US Director of National Intelligence James Clapper included gene editing in his annual “Worldwide Threat Assessment” report to Congress in February 2016. “Given the broad distribution, low cost, and accelerated pace of development of this dual-use technology,” Clapper warned, “its deliberate or unintentional misuse might lead to far-reaching economic and national security implications.”
Experts have put forward several scenarios in which Crispr might be used for nefarious purposes. Some speculate that terrorists might use gene drives to unleash modified pathogens that have been altered to enhance their lethality or make them more infectious. Others have discussed the possibility that terrorists might add a toxin-making gene in the saliva of malaria-transmitting mosquitoes, thus allowing the transfer not only of the disease but the deadly toxin as well. Another scenario envisions the alteration of mosquitoes so they expand their natural habitat and spread diseases such as malaria or dengue fever in non-tropical regions. Some even argue that the expertise required to develop gene drives can be acquired relatively quickly, which would allow amateur scientists or biohackers to develop gene drives at home, posing a more difficult bioterrorism threat to detect and thwart, and creating another reason to be concerned about accidental releases.
In principle, the self-propagating nature of gene drives makes them an ideal tool in the hands of would-be bioterrorists. However, to accurately evaluate their potential misuse, one needs to rigorously assess the state of the technology and consider its limitations. Current fears (and hopes) related to gene drives are based on projections of what gene drives could in theory do if they spread in nature. At the moment, these are still anecdotal, speculative claims and are not based on in-depth empirical research and analysis. One needs to keep in mind that the techniques under debate are still in their infancy, and in spite of their apparent progress, they may not prove to be as dangerous or promising as expected. Developments in synthetic biology have raised similar concerns, yet their expected perils and promises have so far failed to materialize as originally predicted.
One important limitation of gene drives is that they work only with organisms that reproduce sexually, such as animals, insects, and most plants. They cannot be used to alter a virus or bacteria for use as a weapon. It’s also worth remembering that to produce rapid effects, gene drives require species that have a short reproductive cycle, such as insects. Should a rogue scientist or terrorist succeed in altering the human genome, the alteration would take centuries or more to spread to the whole population.
There are also significant technical challenges that a terrorist or an amateur scientist might not be able to overcome. One of the key difficulties of this new technology is getting the gene drive into the organism. The team that developed a malaria-resistant mosquito indicated that only two males out of more than 25,000 mosquitoes carried the drive. Once in the animal, the drive was transmitted by males to 99 percent of their offspring. Females, however, transmitted the drive only at a rate slightly higher than natural reproduction would yield, which would compromise the success of the drive in the wild. Inserting gene drives in bigger animals such as mice presents similar challenges. In addition, laboratory-raised animals sometimes fare poorly on the sexual front once released in the wild, which would compromise the spread of a gene drive.
Another major challenge is the lack of control scientists have over the actions of the Cas enzyme, the cutting function of the drive, which may cut the target multiple times, eventually damaging it, or cut at the wrong spot. Scientists at Rockefeller University have designed a mechanism that could solve this problem, but it has only been tested on cells in petri dishes; it is not clear that it would work as well when used in a lab animal. The Cas enzyme poses several other technical challenges, and scientists have embarked on a quest to find alternative enzymes better suited for their type of work and selected organisms.
Finally, gene-editing experiments have been conducted thus far by a small community of scientists, with teams that include individuals with different types of expertise. In such teams only a couple of individuals may have the skills to insert the drive in the animal. In addition, laboratory expertise with animals requires specialized skills that may not transfer to work with a different species. The laboratory equipment required is not cheap; scientists estimate costs between $100,000 to $200,000, and some of it is specific to the organism used. And when problems occur, scientists must have the ability to tweak the protocols, a skill acquired only through expert knowledge and substantial experience troubleshooting such challenges in a laboratory environment. In this context, gene drives would seem to be beyond the capabilities of terrorists or biohackers with limited scientific knowledge and skills.
State-level use of Crispr and gene-drive technologies to develop bioweapons capabilities could pose a more serious concern. This is because state-level programs would likely have the skills, expertise, resources, and long-term commitment to work out technical problems and develop a long-term research agenda. Members of the Biological Weapons Convention should therefore be looking at how to address this technological development, as should related arms-control bodies like the Australia Group. This is not to assume, however, that states would have an easy time developing these technologies for harm. We know from in-depth historical research that both the United States and Soviet Union took decades to develop and adapt new techniques for biowarfare—and that there were many social, economic, and political factors that affected the technical work within and across all past bioweapons programs. To date, the large body of empirical work on both state and non-state programs indicates that the process of developing biological weapons, even when incorporating the latest technological advances, is intimately shaped by both social and technical factors, and is fraught with challenges. Merely having access to materials, equipment, and even explicit knowledge is not sufficient—tacit knowledge and solutions to a host of social and organizational issues are also critically important.
A realistic approach. Without a clear and detailed understanding of the range of social and technical factors that cause scientists to succeed or fail in their gene-drive endeavors, threat estimates can only rely on speculation and fantasy rather than fact. This has been a problem with many previous security assessments of emerging biotechnologies and life science experiments, such as work in 1980s and 1990s with genetic engineering, in the 2000s with synthetic biology, and more recently with the gain-of-function experiments surrounding the avian flu virus. All of these new scientific developments were seen at the time as harbingers of an imminent bioweapons threat—yet the worst-case scenarios have yet to be seen. Even the bioterrorist attack predicted by a US government commission in 2008 failed to materialize within the estimated five years—or since. Clearly, there needs to be a better understanding of the complex set of factors that can shape the motivations and capabilities of states and non-state actors alike when it comes to developing bioweapons.
To remedy this problem, we are engaged in a project that aims to understand the social and technical factors for how Crispr scientists around the world actually work in the lab. In addition to making an inventory of Crispr-related research, our project aims to address such questions as: What combination of expertise is required to conduct the laboratory work and solve technical problems? Which challenges are common to all gene-drive experiments, and which are specific to the organism used? And what social and organizational issues help or hurt this work? To date, interviews we have conducted with Crispr scientists suggest that making Crispr-related technologies work in practice can require particular kinds of skills, resources, equipment, and infrastructure that are not accessible to all—making the application of Crispr for bioweapons purposes perhaps not as easy as it might seem at first glance.
We hope that empirical data gathered through this project will foster a more informed and nuanced debate about the benefits and dangers of gene editing and gene drives in which the general public can truly be engaged. This empirical data and public engagement should also help scientists and policy makers craft a smart approach to gene-drive oversight, both at the national and international level, without getting distracted by doomsday predictions. Only through such a fact-based approach can the world make sense of new technologies like Crispr and gene drive in a responsible way.