Prudence and policy - reducing the risks of synthetic biology

In theory synthetic biology is the bioterrorists’ best friend. So, ask Damon Terrill and Andy Peek, if the technology is evolving to increase the opportunities for misuse, what can and ought we do about it?

CONTEMPORARY concern about terrorism has occasioned significant attention both in government and industry to the risks associated with the potential misuse of various technologies, including those related to molecular biology. Synthetic genomics technology has been no exception. Commercial and other producers of synthetic oligonucleotides and genes have taken the lead in implementing screening and other approaches that reduce the likelihood of their products’ misuse while working with the academic, non-profit, and public sectors to study what additional measures might be appropriate in future.

For some time, the J. Craig Venter Institute has been investigating the minimal gene set required to sustain the bacterium Mycoplasma genitalium.1 To do this, they have completely re-constructed the organism’s genome in vitro using gene synthesis techniques, and can now control which genes are included and excluded. Their ultimate objective is to engineer bacteria with novel and useful characteristics by manipulating not only gene expression, but also by choosing the genes that are included in the engineered genome – including those from M. genitalium and some from other organisms. Many others are pursuing similar work and goals.  While developing the technologies, these and other scientists, and the policy-makers considering their implications, have engaged in a robust discussion about how best to avoid the misapplication of the techniques now available (and those on the horizon) for potentially harmful purposes.

The Venter Institute and those like it utilise the finest scientists and most advanced laboratories available. The expertise and infrastructure required to do even rudimentary imitations of their work is almost certainly out of reach for potentially criminal actors, even if such work could be kept clandestine. Is a discussion about how criminals or others could intentionally misuse synthetic genomics technology therefore purely hypothetical, at least today?

In synthetic genomics, as in other fields, technological advancement is rapidly reducing the expense, time, and sophistication required of individual users to produce results unthinkable in the recent past; consider the expertise and equipment needed to produce the full-color photographical record of a typical wedding just 20 years ago, and that can now be acheived, by anyone, with a digital camera and a printer. Scientists can now synthesize larger portions (e.g., 500 bases) of genes in significantly less time and at dramatically lower cost than required for the synthesis of a simple oligonucleotide primer just 15 years ago. A consequence is that it is now as cost-effective and almost certainly takes less time to purchase from a commercial provider a synthetic gene than to isolate it from most organisms. As its techniques become more widely used, progress in synthesis technology is accelerating. As a result, reasonable participants in and observers of synthetic genomic’s evolution have already become concerned that – in a future near enough for which to plan - it may be possible to synthesize entire genomes in time frames short enough and at costs low enough to make the technology an attractive tool to those who would abuse it.

At present, it would still be immeasurably simpler (although not at all straightforward) for a criminal to obtain a whole, viable pathogenic organism from other sources. Anthrax, for example, can be isolated from the wild and could be successfully cultured, but the process is laborious and may yield a strain that is not well suited for “weaponisation.” In most cases, the only sources for pure cultures of pathogens are the few carefully controlled, secure laboratories authorised to work with them.

Given the technical and regulatory barriers to obtaining a whole pathogen from nature or a laboratory, the theoretical concern has arisen that in future criminals or others might view synthesis as a viable alternative means to acquiring a pathogenic organism. As synthesis technology becomes less expensive and requires less expertise to use, some worry, producing such a synthesized pathogen might be accomplished in an otherwise unremarkable molecular biology laboratory, with readily available equipment, and by a person or people with minimal training (and without much oversight or involvement by others).

If concern about the intentional misuse of synthetic biology is real, and if the technology is evolving to increase the opportunities for misuse, what can and ought we do today to reduce that risk and improve our ability to respond to over time?

In the United States, the synthetic biology industry, the university and research communities, and interested non-profit groups have taken the lead to evaluate potentially effective risk reduction approaches that take account of the technology’s rapid development. The most authoritative written description of the approaches from which future policy makers might choose is contained in the Options for Governance2 report issued by the Venter Institute, MIT, and the Center for Strategic and International Studies. The product of more than a year’s consultations between and among some of the most important producers and consumers of synthetic oligonucleotides and genes, as well as with non-profit and government participants, the report identifies the available “intervention points” in the process of research, ordering, production, and shipping at which voluntary or compulsory efforts to reduce the risk of misuse might be effective. Those efforts might be directed at or taken by commercial producers of synthetic oligonucleotides and genes, owners of synthesizers, researchers, and other users. At some stages, voluntary codes of best practice (for example, in screening) might be most appropriate, where elsewhere sponsored certifications of compliance could be more effective. In other circumstances and intervention points, compulsory licensing or other arrangements could be necessary. These and similar ideas are now the subject of active consideration both within and outside government in the United States.

Many of the companies most active in gene synthesis have already implemented some aspects of the protections described in the Options for Governance report, in particular the sequence screening of gene orders and those who place them.  Ordering a gene requires that the customer specify the desired sequence, so that screening the ordered gene against databases of sequences of known pathogens and toxins is possible. If screening yields a significant match to one of those databases, synthesis companies can consult with the prospective customer to confirm the intended use of the gene is appropriate, ensure that it can be synthesized safely, and plan for compliance with potentially applicable safety and other regulations.

Dr Robert Jones of Craic Computing, LLC, introduced a first-generation screening programme freely accessible to interested companies. When designing the programme – called BlackWatch - His approach was straightforward: to compare the sequences of oligonucleotides or entire genes against a curated database of sequences from known pathogens - the so-called “Select Agents” as they are known in U.S. regulations. Using the BLAST sequence comparison software from NCBI,3 if the computer comparing the sequences tags a gene as related sufficiently to a pathogen in the database, it alerts the system’s operator who can study the sequence more closely.

Use of this and similar sequence screening systems by commercial oligonucleotide and gene synthesis companies helped to identify the challenges that more widespread screening would pose, in particular as the use of synthetic genes is set to grow and the industry to expand. Most positive matches are to genes intended for unquestionably legitimate purposes; evolutionary sequence preservation also means that matches are frequently to harmless organisms. “False positives” (and accurate positives in uncontroversial circumstances) could mean expensive and overwhelmingly burdensome analysis and delay for producers, users, and regulators alike if present methods were applied in the context of greatly increased synthesis volumes.

The next generation of screening software is designed to address the false positive problem. Dr Jones’ design screens potential matches against the sequences of “secondary databases” comprised of sequences of organisms closely related to target pathogens. A closer relationship to the sequence of a non-pathogenic organism suggests that an otherwise apparent match may be a false positive. Each sequence apparently related to a pathogenic organism will also be screened to identify regions that are conserved among related species and those that are actually indicative of the pathogen. Dr Jones’ more recent approach allows evaluation of potential matches by their location on the pathogen genome, in addition to their similarity scores, enabling a more precise measure of confidence in each match (and the reduction of false positives).

This more advanced screening software, called “Safeguard,” may help synthesis companies to accommodate larger numbers of orders while reducing the burden on them and their customers associated with screening orders related to legitimate research or to sequences unrelated to pathogens. As with Blackwatch, Dr Jones intends to make Safeguard freely available to synthesis companies and others legitimately interested in screening.

With every scientific and technological transformation has come the potential for abuse. In the case of synthetic biology, industry, government and others have come together to plan for likely future risks. To avoid misuse of the technology both now and as it develops, sound business practices, effective screening software, and sensible regulation may all play a part. Experience with sequence screening software has demonstrated the changes that will be necessary as gene sequencing becomes more widely practiced. Industry, users, and regulators will need to adapt, as well.

References
1: Gibson, D. G., Benders, G. A., Andrews-Pfannkoch, C., Denisova, E. A., Baden-Tillson, H., Zaveri, J., Stockwell, T. B., Brownley, A., Thomas, D. W., Algire, M. A., Merryman, C., Young, L., Noskov, V. N., Glass, J. I., Venter, J. C., Hutchison, C. A. 3rd & Smith, H. O.  (2008) Complete chemical synthesis, assembly, and cloning of a Mycoplasma genitalium genome. Science. 319(5867):1215-20.

2: Garfinkel, M. S., Endy, D., Epstein, G. L. & Friedman, R. M. (2007) Synthetic Genomics – Options for Governance. Joint publication by J. Craig Venter Institute, Massachusetts Institute of Technology and Center for Strategic and International Studies.

3: Altschul, S. F., Madden, T. L., Schäffer, A. A., Zhang, J., Zhang, Z., Miller, W. & Lipman, D. J. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25(17):3389-402.