In his April 2006 report “Uniting against terrorism”, the United Nations Secretary General suggested:

“The approach to fighting the abuse of biotechnology for terrorist purposes will have more in common with measures against cybercrime than with the work to control nuclear proliferation.”

This, and statements like it, have become common currency where folks gather to talk about how to deal with the potential that the life sciences or biotechnology might be misused to cause deliberate harm.

More broadly, because of the nature of biotechnology, measures developed to limit the spread of nuclear weapons are not a good fit. Nuclear weapons technology includes many single use items (no one has been able to give me a peaceful use for Uranium enriched beyond about 80%); biotechnology, by contrast, is almost all dual use. Nuclear technology is in the hands of, or very closely regulated by, governments. There are also comparatively few users and they are focused at a limited number of facilities. Biotechnology is in the hands of the private sector - and increasingly in the hands of the public directly. It is used by wide variety of actors, in a distributed manner.

A case in point would be the control list - a core element of efforts to prevent technology being used as a weapon. By identifying a specific set of items or technologies that pose specific risks of being misused to cause deliberate harm, countries attempt to limit the impact of controls to only those items they believe to pose the greatest risk. Such lists are then used at the international level to control transfers of certain items and at the national level to regulate access to them. In a biological context, they all include lists of biological agents, described by their taxonomic name. They are in use throughout the world (from China to the United States, and Europe to India) and are under development in many other countries (for example, Malaysia reported to the 2014 BWC Meeting of States Parties on its efforts to develop such a list).

Taxonomy-based lists cover things that do not need to be covered and miss some of those that need to be. Basing control lists on an agents name (or taxonomy) is a mismatch to risk. The risk comes from a combination of biological properties. Agents are on these lists because, for example, they produce a toxin, have particularly environmental properties, or the ability to infect certain cell types, etc.. It has long been established that biological function is not a correlate of taxonomy. There are agents that share the same name but do not share functional properties. Equally there are also related species with different names that have functionally identical properties. In 2006, Richard Okinaka et al demonstrated all this for the causative organism for anthrax.

Further, taxonomy-based lists make it difficult to anticipate the changing technology landscape. They tend to be reactive. For example, none of the control lists mentioned in this post cover the potential application of genome editing technology and gene drives recently highlighted as being potentially able to make all mosquitos resistant to the malarial parasite in a single season. This technology was covered in science and technologies reviews under the BWC (for example, the CRISPR/CAS9 technology was discussed by the ISU and Switzerland and Gene drives were covered by Professor Ken Oye in a guest presentation.) In this example, none of the parts used came from a listed pathogen but in theory, the same tools could be used to very rapidly increase the spread of vector borne diseases, massively increasing the impact of a biological attack.

Because of bad fit taxonomy-based approaches vs the risks that governments are actually concerned about, lists using them are often difficult to interpret at the practical level. That also makes them difficult to scale. By contrast, focusing on the function of biological material (rather than where it comes from) holds many benefits. In particular, it makes it easier to develop tools, guidance and standard operating procedures that simplify implementation, thereby reducing burden and making such efforts more scalable. The experiences of the Safety Committee at the International Genetically Engineered Machines competition (iGEM) illustrates that significant strides are being made towards partial functional screening.  Spaces like iGEM provide very useful testbeds for further improving and realising functional screening approaches.

Three trends in the development of science and technology will compound difficulties in using taxonomy-based approaches:

Distribution of technology

Many countries around the world are investing heavily in biotechnology. For example, Malaysia reported to the BWC last December that it envisages that by 2020 the bioeconomy “will employ up to 160,000 people, contributing to 5% of the nation’s GDP, and is expected to generate RM248 billion in revenue”. Biology will be a major manufacturing technology. It’s spread around the world is going to be rapid.

Distribution of users

Science is being done in places outside of traditional settings (for example in community labs like Genspace in New York). Some of these spaces have little history of regulatory compliance. Others users are conceptually opposed to government oversight. Some users, such as bioincubators, require streamlined oversight if there are to deliver their raison d'etre. Both types of user are increasingly global, for example with DIYBio labs in Asia, and bioincubators in the UAE.

Distribution of process

Developing the bioeconomy is an international exercise. For example, the development of semi-synthetic artemisinin, as well as other uses of the platform to produce biofuels and fine chemicals illustrates that it is increasingly common that demand in one country results in research in a second, that is scaled up using commercial partners in a third, before being produced in a fourth country.

In conclusion, it is time for efforts to deal with the potential for biotechnology to be used to make weapons to start moving past taxonomy-based approaches and exploring how to target directly those biological properties that cause the concerns. Whilst taxonomy-based approaches have served us well whilst we have had no real alternative, that will not be the case for long. There is a need for research and development to speed up efforts to make function-based screening a reality. It is one key to being able to unlock the full potential of biology as a manufacturing technology by allowing more flexible, anticipatory and scalable efforts to safeguard the bioeconomy.

---

The author wishes to thank Dr. Ken Oye and Dr. Todd Kuiken for their collaboration on gathering the concepts in this post. It draws heavily on comments prepared for the US Government in response to requests for perspectives on the implementation of the Select Agent Rules. Similar material was included in the Biosecure presentation on developments in science and technology at the academic portion of the event to mark the 40th Anniversary of the Biological Weapons Convention, held in the United Nations Office at Geneva on Monday 30 March 2015.