Strengthening Security for Gene Synthesis: Recommendations for Governance

Abstract

Since the inception of gene synthesis technologies, there have been concerns about possible misuse. Using gene synthesis, pathogens—particularly small viruses—may be assembled “from scratch” in the laboratory, evading the regulatory regimes many nations have in place to control unauthorized access to dangerous pathogens. Progress has been made to reduce these risks. In 2010, the US Department of Health and Human Services (HHS) published guidance for commercial gene synthesis providers that included sequence screening of the orders and customer screening. The industry-led International Gene Synthesis Consortium (IGSC) was formed in 2009 to share sequence and customer screening methods, and it now includes the major international gene synthesis providers among its members. Since the 2010 HHS Guidance was released, however, there have been changes in gene synthesis technologies and market conditions that have reduced the efficacy of these biosecurity protections, leading to questions about whether the 2010 HHS Guidance should be updated, what changes could make it more effective, and what other international governance efforts could be undertaken to reduce the risks of misuse of gene synthesis products. This article describes these conditions and recommends actions that governments should take to reduce these risks and engage other nations involved in gene synthesis research.

Since the inception of gene synthesis technologies, there have been concerns about possible misuse. Using gene synthesis, pathogens can be assembled “from scratch” in the laboratory, evading the regulatory regimes many nations have in place. Progress has been made to reduce these risks, but there have been changes in gene synthesis technologies and market conditions that have reduced the efficacy of the existing biosecurity protections. This article describes these conditions and recommends actions that governments should take to reduce these risks and engage other nations involved in gene synthesis research.

Since the 1970s, scientists have performed gene synthesis, the process of designing and synthesizing sequences of double stranded DNA (dsDNA).¹ Gene synthesis allows researchers to design and create a gene “from scratch,” without an existing DNA template. As gene synthesis technologies have become more accurate, affordable, and able to generate progressively longer strands of DNA, life science researchers have increasingly relied on them to create advanced gene-editing systems, to identify new drug targets, and even to explore DNA-based data storage.^2-4 These products are used as research tools in laboratories all over the world, whether they are small oligonucleotides or “oligos,” pieces of DNA from 15 to 200 bases in length, or longer genes thousands of bases in length.^5-7 Oligonucleotides are an essential tool for researchers working with DNA or RNA to be able to learn the genetic sequence or to help quantify gene expression. Some laboratory researchers have DNA synthesizers in their laboratories to make oligos. These synthesizers can cost between US$10,000 and US$12,000 for a new model. Other scientists may use commercial gene synthesis providers, particularly for more complex, longer fragments that require special expertise or purification steps. The global gene synthesis industry, including DNA synthesis, sequencing, and assembly, genome editing, and bioinformatics, was worth about $217 million in 2017 and is projected to be over $600 million by 2022.^8,9

Gene synthesis technologies produce numerous benefits—novel drug targets, vaccine development, and gene therapy approaches among them—but they also raise biosecurity concerns about misuse. The most direct biosecurity concern is that they could be used to make a pathogen from scratch. Using gene synthesis tools, a nefarious actor would not need access to a pathogen (eg, from a laboratory sample) to be able to acquire it, make more, and develop it into a weapon. Acquiring the genetic material that encodes a pathogen is the first step in being able to, in the parlance of synthetic biology, “boot up” an infectious pathogen. The ability to boot up DNA encoding viral genomes has been repeatedly demonstrated in the laboratory with a wide variety of viruses: with poliovirus, 1918 flu, Ebola, and horsepox, a close cousin to smallpox virus.^10-13 As biosecurity controls in the United States and many other nations are primarily based on pathogen access, such as through the US Federal Select Agent Program, gene synthesis technologies undercut these protections.¹⁴

While samples of most regulated pathogens, including viruses as virulent as Ebola, can be found in the natural world or during outbreaks, access to samples of the eradicated pathogens rinderpest virus and smallpox (variola) are much more restricted. However, the genetic sequence information for most pathogens is freely available—for example, from the National Center for Biotechnology Information (NCBI), part of the US National Institutes of Health (NIH). These sequences have been quite important as a research tool, but their availability leaves open the possibility that known, dangerous pathogens could be made by people who intend to use them for harm.¹⁵

Even if a researcher already has access to a pathogen, gene synthesis technologies could allow nefarious actors to introduce further changes to increase its pathogenicity or transmissibility or to take advantage of beneficial disease research. For example, scientific researchers are exploring designed oligonucleotides that bind to and block the action of noncoding RNAs as novel drug targets for certain conditions such as Duchenne muscular dystrophy or some cancers.¹⁶ However, this same approach could also be used to design oligonucleotides that bind to and block RNA action crucial to healthy cell survival.

The de novo synthesis of known pathogens, particularly small viruses, is listed as one of the most pressing biodefense risks according to the 2018 National Academies of Sciences report Biodefense in the Age of Synthetic Biology.¹⁷ At this time, concerns about misuse of gene synthesis to make entire pathogens from scratch are almost entirely limited to viruses. Synthesis of whole cellular genomes, bacterial or fungal, is a much more challenging task that has only been accomplished by a few groups.¹⁸ This research is limited by technical challenges; whole cellular chromosomes are large, fragile molecules and, unlike viruses, typically do not have a built-in molecular infection mechanism that can be co-opted to easily render them functional within a cell.

That gene synthesis technologies could be misused was discussed at the earliest synthetic biology scientific meetings.^19,20 The lead scientist in the 2002 synthesis of polio, Eckard Wimmer, referred to his work as a “wake-up call” that should lead providers of DNA synthesis products to screen their orders.²¹ The security vulnerabilities in gene synthesis became even more clear to the public when a reporter from the Guardian, a UK-based newspaper, placed an online order to a gene-synthesis biotechnology company for a piece of smallpox DNA in 2006.²² The order should have looked suspicious: It was to be delivered to a residential address, not a laboratory; the company name was made up; and the sequence ordered was for an eradicated pathogen. Nonetheless, the order was filled, and the DNA was successfully procured by the reporter. The complete smallpox virus is 52,000 bases, much larger than the 78-base length the reporter ordered, and there was no danger to the public, but the point was made: Gene synthesis companies could unwittingly provide services for nefarious actors. Software to screen orders was developed, major gene synthesis providers adopted screening protocols, and options for governance of gene synthesis were debated by policy analysts and scientists.^23,24

2010 HHS Guidance and Provider Responses

In 2010, the US Department of Health and Human Services (HHS) published guidance for commercial gene synthesis providers that included sequence screening of the orders (focusing on whether the gene synthesis product could be used to create a dangerous pathogen or be subject to misuse) and customer screening (focusing on whether the customer is authorized to receive the gene synthesis products).²⁵ This voluntary HHS “Screening Framework Guidance for Providers of Synthetic Double-Stranded DNA” makes specific recommendations to dsDNA providers and also reminds providers of their obligations under existing federal regulations, including the Select Agent Regulations, Export Administration Regulations (EAR), and criminal statutes. To prevent unauthorized users from obtaining certain regulated materials, the guidance states that dsDNA providers have 2 “overriding responsibilities”: They “should know to whom they are distributing a product,” and they “should know if the product that they are synthesizing and distributing contains, in part or in whole, a ‘sequence of concern.’ ”²⁵ The guidance recommends customer and sequence screening for orders of more than 200 base pairs (bp).

The authors of the guidance made the decision not to recommend a reference database for screening, because “generating a comprehensive list of such agents to screen against is not currently feasible” and because databases used for research, such as GenBank, are added to in the course of research and are up to date.²⁵ Thus, the guidance does not specify which database should be used for screening but holds as an example GenBank, the NCBI NIH database, one of several available annotated collections of publicly available DNA sequences.²⁶ Identifying an order as a “sequence of concern” does not necessarily mean that it cannot be filled, but it should trigger follow-up screening to determine if other regulatory barriers should prevent fulfillment. The guidance recognizes that there are many portions of the genome of regulated pathogens that are not pathogen-specific (eg, so-called “housekeeping” genes, variations of which are found in many nonregulated organisms and viruses), so the sequence of concern should ideally be unique to a regulated pathogen.

Providers of dsDNA are guided to perform a minimum sequence screening against the Select Agents and Toxins list for all orders and additionally against the Commerce Control List (CCL) for international orders. Screening against the CCL brings providers into compliance with the Export Administration Regulations (EAR) and with US membership obligations to the Australia Group, which is a multilateral export harmonization regime.^25,27 For customer screening, the guidance recommends that providers screen all orders against lists of prohibited persons and entities, including the Department of Treasury Office of Foreign Assets Control (OFAC) list of Specially Designated Nationals and Blocked Persons (SDN List), the Department of State list of individuals engaged in proliferation activities, and the Department of Commerce Denied Persons List (DPL). dsDNA providers are also advised in the guidance to retain all order records for at least 8 years according to criminal statutes.

The 2010 HHS guidance formalized what many gene synthesis providers were already doing: screening their customers' orders for sequences encoding dangerous pathogens. The International Gene Synthesis Consortium (IGSC) was formed in 2009 as an industry-led, international, collaborative group of gene synthesis companies to strengthen biosecurity.²⁸ The IGSC published their own Harmonized Screening Protocol to provide details on how companies can implement and incorporate the 2010 HHS Guidance into their business practices.⁴ According to their website, IGSC member-companies constitute approximately 80% of worldwide gene synthesis capabilities. Members include Ginkgo Bioworks, Integrated DNA Technologies (IDT), BGI, Blue Heron, ATUM, ThermoFisher Scientific, GenScript, SGI-DNA, Twist Bioscience, Bioneer Corp, Edinburgh Genome Foundry, and Battelle Memorial Institute.

Aims and Methods

Since the 2010 HHS Guidance was released, there have been dramatic changes in gene synthesis technologies and market conditions that have reduced the efficacy of these biosecurity protections. There are many more companies worldwide that provide gene synthesis products to order, and laboratory techniques have been developed that allow for the assembly of genomes from smaller DNA pieces. This context leads to questions about whether the 2010 HHS Guidance should be updated, what changes could make it more effective, and what other efforts outside of the HHS Guidance could be undertaken to reduce the risks of misuse of gene synthesis products.

To address these questions, the authors discussed options for governance with users and providers of gene synthesis products, as well as longtime analysts in the field. With support from the Open Philanthropy Project, the Johns Hopkins Center for Health Security also hosted a 1-day meeting on April 9, 2019, in Washington, DC, to discuss the current guidelines and gaps in governance that could be addressed.²⁹ Invited speakers and audience members discussed potential steps that the US government and providers could take to increase biosecurity in gene synthesis. Representatives from HHS took part in the meeting, as well as other stakeholders from government, academia, and industry. The authors derived a series of recommendations based on an analysis of the literature, an assessment of the current gene synthesis landscape, interviews with experts in the field, and input from the April meeting with industry, government, and researcher participants. Here we describe limitations and challenges to current approaches to gene synthesis governance and recommend actions that governments should take to reduce these risks.

New Approaches to Gene Assembly

Since 2010, there have been technical advances that challenge or evade the biosecurity benefits of gene synthesis screening protocols. It is now more straightforward to assemble large pieces of genetic material using methods other than purchasing screened DNA synthesis products. While the ability to “read” (sequence) DNA is still more rapid and accurate than the ability to “write” (synthesize) DNA, new technologies are closing that gap, which also challenges biosecurity screening controls.³ Some of the most important advances that diminish the effectiveness of current gene synthesis screening approaches are Gibson Assembly, enzymatic assembly of DNA, genetic recoding, CRISPR, and a new type of desktop DNA synthesizer, a product that is just on the horizon. This section of the report describes these technical advances and the challenges they pose to current gene synthesis screening protocols.

Gibson Assembly is a widely used synthetic biology technique that can be used to rapidly and accurately assemble large genetic fragments from oligonucleotide fragments or from single-stranded or double-stranded DNA oligonucleotides. Using Gibson Assembly, smaller pieces of DNA (which are now unscreened) may be assembled to construct much larger fragments.³⁰ When oligonucleotides are used as inputs to the Gibson Assembly reaction, they are typically 60 bp, but they can be as small as 40 bp, and the largest product that Gibson Assembly can produce is in the order of 200 kb. Gibson Assembly uses DNA replication enzymes (Fusion DNA polymerase, a 5′ endonuclease such as T7, and TAQ ligase) to merge smaller DNA fragments into 1 large construct by leveraging regions of overlapping DNA.³⁰ The advantages of Gibson Assembly lie in its simplicity and ease of use: It does not require expensive equipment such as a thermocycler; 2 to 15 fragments can be combined simultaneously, allowing for large pieces of synthetic DNA to be constructed in a single tube; and it is compatible with high-throughput applications.³¹ Since it was invented by Gibson et al in 2009, the process has become standard procedure in both academic and industrial laboratories, and it has undergone improvement for higher efficiency and accuracy. Previous methods relied on finding unique sites for restriction enzymes within the sequence to be constructed and which could produce compatible ends for assembly, often based on overlaps no larger than 3 to 4 bases. Gibson Assembly is less labor intensive and less susceptible to error.³²

Gibson Assembly is perhaps the most popular method of making large DNA constructs out of smaller pieces, but it is not the only one: There are several “scarless assembly” methods (scarless because they do not require restriction enzymes to assemble), and it is very likely that there will be further technical improvements in the years ahead. Scarless methods include SLIC, CPEC, SLiCE, and Yeast Spheroplast Assembly, and there are older methods that still may be used, including Golden Gate and PCA.^18,33-36 SLIC and CPEC use purified enzyme biochemistries in vitro, similar to Gibson Assembly.^34,35 SLiCE is in vitro but instead of purified enzymes uses lysates of a DH10B Escherichia coli strain that is optimized to express a λ prophage Red recombination system.³⁵ Yeast Spheroplast Assembly operates in vivo using living yeast cells transformed with the input DNA fragments and with the output of a fully assembled DNA product cloned into the resulting transformed yeast colonies. Yeast Spheroplast Assembly allows for larger DNA constructs to be assembled and cloned (up to 1 megabase) compared to any of the other scarless systems (typically on the order of 100 to 200 kb).³⁶ Most of these assembly systems use purified, hard to produce enzymes that laboratories typically purchase from biological supply companies, but SLiCE and Yeast Spheroplast Assembly use easy-to-culture, commonly used living organisms as reagents. Thus, SLiCE and Yeast Spheroplast Assembly are difficult to regulate from a product control perspective.

Enzymatic synthesis, or “nontemplated polymerase synthesis,” is another technology that challenges the biosecurity protections afforded by DNA synthesis screening. So far, all large DNA synthesis has been built on the concept of stitching together small pieces of DNA to build bigger ones (initially with ligase and restriction enzymes, and later with the scarless methods described above). The pieces of DNA to be stitched together ultimately came from oligonucleotides, which have been produced the same way for the past 30 years: with organic chemistry methods and phosphoramidite precursors. Enzymatic synthesis is meant to replace that phosphoramidite chemistry step by using the activity of the enzyme terminal deoxynucleotidyl transferase (TdT).³⁷ TdT, unlike most DNA polymerases, does not build DNA of a particular sequence (the growing strand) based on the complementary sequence of another template strand of DNA, but rather randomly adds nucleotides from solution to a growing strand. If allowed to do so, TdT will add thousands of such nucleotides in vitro; it has been demonstrated that this process can be controlled by providing nucleotides of a desired identity into the reaction solution (either A, T, G, or C) and chemically blocking them so only one is incorporated at a time.³⁷ Using this technique, researchers have direct control over the exact sequence of a growing strand of DNA. Currently, the demonstrated length is only 10 bases in length, nowhere near competing with conventional oligo products on the market. This technology is being commercialized by Molecular Assemblies.³⁸ While still early in development, this technology has the potential to dramatically reduce the cost of synthetic DNA and blur the lines between products like oligos and larger synthetic DNA constructs.

Genetic Recoding

Another technical development that affects synthesis screening is genetic recoding, because the sequence ordered by a customer may be inscrutable to the gene synthesis provider. Scientists have been expanding the language of DNA (A, T, G, C) by adding in new bases (S, B, P, and Z). These non-natural, synthetic bases expand genetic coding possibilities while closely resembling the natural bases in chemical structure and in how they bond to and interact with each other.³⁹ Although synthesis of these “unnatural” bases was first demonstrated in the 1980s by Piccirilli et al,⁴⁰ it was not until 2014 that they were successfully inserted into functional, living E. coli cells.⁴¹ With this proof of concept, researchers were able to create synthetic complementary bases that function like natural bases, including the way they bind to each other, how they hold their form and information, and how they maintain the integrity of this information in a double helix structure.^42,43 Researchers more recently demonstrated that these unnatural bases can be incorporated into nucleotide triphosphates and into proteins, demonstrating their feasibility to influence an organism and function properly.⁴⁴ With 4 non-natural bases (S, B, P, Z), in addition to the natural bases (A, T, G, C), genetic coding capabilities have been increased greatly, allowing more bioengineering possibilities. These novel bases represent a new frontier for screening regulations, as typical sequence screening would not detect altered codon usage or novel nucleotides. This could reduce the sensitivity of a screening technology and could increase the burden of follow-up screening on the provider's part. Without context and communication with the researcher, the company may think that it is an error or a potential threat, when in fact the researcher may be using these novel nucleotides in a completely safe manner. Use of novel codons and nucleotides warrants specific guidelines in screening for the benefit of the researcher and the provider.

Researchers were recently able to recode the entire E. coli bacterial genome using synonymous codons, resulting in a viable organism.⁴⁵ Such synthetic strains of E. coli may have countless applications, but one example is its use in the production of synthetic polymers by reassigning the amino acids of existing codons. In Fredens et al, the REXER method allowed researchers to introduce large (100 kb) synthetic DNA strands into the E. coli genome, then “reassign” the synonymous codon of a natural amino acid with a synthetic one.⁴⁵ The REXER method essentially allowed them to explore how extensively the genome could be recoded and opens up more opportunities for future recoding experiments. With or without the addition of non-natural bases, it may soon become possible to recode entire genomes using synthetic codons or using codons of non-natural amino-acid pairings. Recoded genome experiments could be undertaken either to alternate natural amino acids and non-natural amino acids or to incorporate wholly unnatural amino acids, which may be useful for a number of purposes important to biosafety and research.

However, whether or how these non-natural codons and bases would be screened for biosecurity concerns is an open question, as it would require the cooperation and active involvement of researchers using such codons. While DNA incorporating non-natural codons would likely be identified in a sequence screen, the function of the resulting DNA, and the ultimate functional product, would be contingent on an unknown context, which would not be standard across all cellular strains or species. This opens the heretofore unrealized potential for DNA to be ordered as a customer-encrypted version of a dsDNA product that the company and its screening methods cannot decipher. The HHS Guidance recommends that providers of dsDNA screen all orders against sequence databases representing the Select Agents and Toxins list; however, the screening against these genetic databases presumes that the query sequence and the database sequences are genetically coded in the same way.

How Genome Editing Changes Risks

One of the most defining biotechnology advances in recent years is genome editing with CRISPR, which also has consequences for gene synthesis screening. CRISPR provides additional pathways for a nefarious actor to modify organisms to make them more pathogenic, using small pieces of DNA that would not be screened. CRISPR substantially reduces the effort needed to make sequence-targeted enzymatic effects on a DNA or RNA template and could allow for an increase in virulence of an existing organism, which was a top-tier concern of the National Academies of Science and Medicine in their 2018 report.¹⁷ This research has already been done for legitimate research purposes.⁴⁶ CRISPR could allow nefarious actors to engineer pathogens to become more dangerous, or to engineer previously nonpathogenic organisms. New traits, including antibiotic resistance, could be added to an organism. This could occur without using any more synthetic DNA than small oligonucleotide-length pieces, which are not likely to be screened.

Synthesis in Desktop Machines Is Not Screened

Scientists can also perform DNA synthesis in their own facilities, rather than ordering DNA synthesis products from a company. This presents a challenge, since DNA synthesis in a desktop synthesizer involves no screening at this point. There are 2 major types of desktop synthesizers: oligo synthesizers (which have been around for decades, and which are relatively commonly available as used laboratory equipment) and gene synthesizers (a new technology that is not very common). The MerMade oligonucleotide synthesizers allow researchers to directly design and create up to 192 oligos in a few hours.⁴⁷ The design-to-create steps involve no outside companies, apart from the nucleotides that may be ordered, which can be small.

Gene synthesizers, such as the BioXP 3200, are much more advanced and can create DNA libraries with lengths ranging from 400 to 1,800 bases. These synthesizers differ from more traditional DNA synthesizer instruments in that they are more compact, require fewer reagents, and have a more rapid turnaround time.⁴⁸ The BioXP system requires users to design and order oligos with the company, SGI-DNA, and then they may use these to synthesize the libraries; as SGI-DNA is an IGSC member, these orders are likely to be screened. While these gene synthesizers are still error prone and may not be best for large-scale or commercial uses, as these types of synthesizers gain popularity and become more widespread, their secure use should be proactively discussed.⁴⁹ There may be software solutions that could help in screening. Gene synthesizers may allow some screening at the oligo design level, but only if sequences below 200 bp are screened. Oligo synthesizers, however, may require software that performs screening as the oligos are designed, which could theoretically flag and record any hits of sequences of concern, alerting a responsible official.

Changes to the dsDNA Synthesis Market

In addition to the technological advances, the gene synthesis market has also changed since the 2010 HHS Guidance, which affects biosecurity. Costs have decreased, making gene synthesis products more accessible and widespread all over the world. Over the course of 2 decades, the cost of synthesis has fallen from hundreds of US dollars to fractions of US cents per base.⁵⁰ For example, according to the Gene Universal website, their gene synthesis services start at only $0.09 per base with a current promotion, and Genscript services start at $0.11 per base.^51,52

The global value of the DNA synthesis market grew from US$46.3 million in 2010 to US$203 million in 2017, with even more growth projected in the next 5 years, up to US$825 million.⁸ The 4-fold increase in market value of DNA synthesis coupled with a 10-fold decrease in cost of synthesis per base has greatly enabled the growth of both DNA synthesis and the synthetic biology industry as a whole. Raw materials such as oligonucleotides and gene fragments serve as prerequisite materials for important research and development in health care, drug development, and biofuel creation, among other applications.⁸ The global landscape of gene synthesis providers has grown since 2010.⁵³ In 2010, the HHS Guidance listed approximately 45 companies with gene synthesis capabilities. According to recent market research, more than 320 companies are now relevant to the DNA synthesis field.⁵⁴

While the cost of synthesis continues to fall, the cost of biosecurity screening has remained relatively stable, which makes the relative cost of screening to a gene synthesis provider increasingly higher.⁴ Most customer orders can be screened automatically. In the case of a positive hit, however, it is suggested that the provider conduct “follow-up screening,” which is more expensive, as it takes time.²⁵ Follow-up screening may involve referencing the customer's address, affiliation with universities or companies, and past gene synthesis orders. The guidance encourages companies that are unsure of next steps in follow-up screening to contact their local FBI Field Office Weapons of Mass Destruction Coordinator for additional assistance. The primary cost of screening a sequence, regardless of length, is in human analyst time in the event of a positive sequence match to a threat-list sequence. While screening has traditionally been costly to perform, partially due to the cost of screening programs, more open-source screening tools have become available to relieve some of this burden.⁴ The democratization of screening tools has helped keep the screening recommendations feasible as the scale and cost of dsDNA synthesis continue to change.

In 2015, the J. Craig Venter Institute (JCVI) released a report detailing the impact of the 2010 guidelines and how the technology around gene synthesis is changing.⁵⁵ They found that, although only 5% of orders to IGSC companies may appear as positive hits, the cost of investigating these is exorbitant for most companies. Of this 5% of hits, the IGSC further categorizes them into “yellow” and “red” hits, depending on the level of homology with regulated pathogens. Yellow hits have greater than 80% homology to sequences in regulated pathogens, but those sequences are not thought to be related to pathogenicity.⁵⁵ Examples of yellow hits could include housekeeping genes in regulated pathogens, which are specifically excluded from screening requirements in the 2010 HHS Guidance but are still investigated under the IGSC Harmonized Screening Protocol as a best practice. Red hits have greater than 80% homology to sequences in regulated pathogens that do have established relationships with pathogenicity. Yellow hits make up about 4.3% of all submitted orders, while red hits make up about 0.7% of all submitted orders.⁵⁵

Even with this low rate of flagged orders, the cost to dsDNA providers to screen and follow up on these orders will become increasingly burdensome as the profit per base falls. To make up for the decrease in cost per base, companies will have to accept, and therefore screen, more orders. Compared to the time required for customer follow-up, the time required for sequence screening is relatively small—on the order of minutes. Red hits can take several hours to resolve during the customer follow-up phase, because the information needed to verify and then complete these orders cannot be gleaned from a database but rather must be gathered from the customer.⁵⁵ Thus, the customer screening and follow-up component of biosecurity controls for the dsDNA provider will continue to represent a nontrivial burden on overhead costs of gene synthesis.

Additionally, the JCVI report proposed that a central database of sequences of concern, for all provider companies, may help reduce costs of screening by reducing the follow-up time required for researching the positive hit.⁵⁵ Further, a central database or screening software that is capable of placing even fewer orders in the red or yellow categories by making more informed calls on pathogenicity or potential biosecurity threat would be valuable additions to the gene synthesis workflow.

Screening software has grown alongside the synthetic DNA market, with Battelle Memorial Institute's recent ThreatSeq being the most well-known new example of this. ThreatSeq is a screening software coupled with a curated database that aims to improve users' abilities to screen sequences of interest and get meaningful data back.⁵⁶ Battelle employs red team evaluations to challenge the usability and security of the software, aims to provide a “road map” for users, and wants to encourage widespread use of its software. Battelle reports that one of the important characteristics of ThreatSeq is the level of detail in the metadata of each gene, which requires ongoing curation and updates as the field changes. They report that the software covers 70% of human and zoonotic agents. A major synthetic DNA company, Twist Biosciences, recently worked with Battelle to use ThreatSeq and test its abilities.⁵⁷ The company then adopted the software for screening their millions of synthesized sequences.⁵⁷

Aside from industry solutions to improve sequence screening, the US government has funded efforts to address biosecurity vulnerabilities in gene synthesis and gene editing. The Intelligence Advanced Research Projects Activity (IARPA) Functional Genomic and Computational Assessment of Threats (FunGCAT) program aims to improve gene synthesis screening to alert providers to sequences of concern; Finding Engineering-Linked Indicators (FELIX) aims to use novel engineering tools to identify genetically modified organisms in wild habitats; and the Defense Advanced Research Projects Agency (DARPA) Safe Genes project aims to provide therapeutic and preventative treatments to counter unwanted gene editing.^58-60

Third-Party Fulfillment

There is also a new category of gene synthesis providers since 2010: the third-party fulfillment company, which can outsource the synthesis of oligonucleotides, or entire genes, for their customers (often universities, companies, or individuals that do not have a synthesizer on site).⁶¹ Further, third-party fulfillment services fill and complete dsDNA synthesis orders but are not themselves the end users of the product. While these companies primarily deliver oligos to their customers, they can also contract with larger gene synthesis companies, some of whom are in the IGSC, to provide longer gene fragments.⁶² Some third-party services have their own DNA synthesizers, but these capabilities are generally related to filling orders for oligos 100 bases or fewer. Companies providing these services commonly state that the entire process of creation, synthesis, and purification of small oligos can be performed in only 1 to 2 business days.^61,63 Users might choose to employ these services due to the rapid turnaround time and the offer of additional services, such as vector and plasmid optimization, oligo purification, and oligo desalting.⁶⁴

Ultimately, these types of fulfillment services mask the identity of the end user to the actual synthesizers of the gene fragments or oligos. It is unclear what responsibilities these third-party companies have for customer or sequence screening, and the end user of the product may also not be screened either by them or by the contracted dsDNA provider. In the customer verification section of the guidance, it is recommended that dsDNA providers verify each customer's full name and contact information, billing and shipping addresses, and institutional or corporate affiliation.²⁵ Indeed, some of these third-party services state that sequence screening occurs for each order, but it is still unclear by whom the screening is performed.⁶⁵ Clearer recommendations for screening responsibilities between third-party fulfillment companies and providers of dsDNA products may help to close these information gaps. At a minimum, the identity of the end user could be made available to the providers of dsDNA products in order to complete their own critical customer screening functions.

Increasing Biosecurity Controls on Gene Synthesis

Even in 2010, when the HHS Guidance was released, it was a partial solution to biosecurity concerns, as it was possible even then for actors to acquire gene synthesis products by circumventing the guidance, and possible to acquire dangerous pathogens by other means. Commercial providers of more than 200-bp gene synthesis products were not the only source for the genetic material that encodes dangerous pathogens. Other methods, such as using laboratory synthesizers, could yield these products, although perhaps with more impurities and requiring more patience. Gene synthesis technologies were international, not all companies complied with it, and there was no universal international standard. And many dangerous pathogens can be found in laboratory stocks or in nature, where they cause infections in people and animals; if a pathogen can be acquired in those settings, there would be no need to go to the trouble of synthesizing it in the laboratory.

There are now even more challenges to the protections provided by gene synthesis screening. There are additional technical routes available to produce longer genes and full-size genomes, as demonstrated by the synthesis of a minimal bacterial genome by Gibson Assembly in 2010.⁶⁶ Additionally, the number of commercial providers globally has increased, and not all providers screen ordered sequences or perform customer screening; desktop synthesis technologies are becoming closer to market; and there are financial pressures on companies that discourage taking on the burden of screening.

Despite these challenges, we think that there are good reasons to both reinforce and expand the norms to the extent possible around gene synthesis screening, recognizing that screening will only ever be a partial solution to biosecurity concerns. There are several reasons to work to strengthen the approach to synthesis screening. First, gene synthesis screening sets a standard of responsible behavior for commercial providers of genetic material.⁶⁷ Second, it may be a useful tool for biosafety if it prevents imprudent and unsafe ordering of genes from dangerous pathogens without due consideration of risks. And third, it may deter or detect certain types of nefarious actors. State-sponsored actors are unlikely to be detected or deterred by gene synthesis screening controls, given that they would presumably have their own capacities for gene synthesis, but individuals or nonstate actors might. It remains true now, just as when the Guardian published their exposé: People should not be able to easily order the DNA encoding smallpox from the internet.

Expanding the norms around gene synthesis is also important because of the nature of DNA research: It is ubiquitous. DNA research is accessible to everyone, from elementary school students to DIY biologists to tenured professors in a range of scientific disciplines. CRISPR and other advances have helped to make the tools needed to work with DNA cheaper, and extensive research in the scientific community provides a wealth of information for those interested in working with DNA. Further, as the cost of synthetic DNA has steadily decreased in recent years, this type of research is accessible to researchers in many different environments. This underscores the difference between working with synthetic DNA and whole organisms. While DNA may be manipulated on its own, or even transformed into E. coli or yeast, the reagents and organisms needed are easy to find, relatively safe, and inexpensive. Even in resource-poor settings, one can use synthetic DNA to better understand the function of a gene or attempt to manipulate gene expression. This is in contrast to other types of research, which remain resource intensive, challenging, and slow-going.

Recommendations for Governments

Given the importance of preventing the nefarious synthesis of pathogens, there are steps that can and should be taken, even in the face of advancing technologies and technologies that may evade screening. These steps should be taken alongside other kinds of biosafety and biosecurity controls intended to prevent, deter, detect, attribute, and mitigate misuse of life science tools. The United States and other governments should build on the foundation of the 2010 HHS Guidance and innovate their guidance to accommodate the growing gene synthesis industry.

Governments should institute requirements for their life science research grantees that gene synthesis products be purchased from companies that screen. In the 2010 Guidance, there is no mention of the purchasers of DNA, only the gene synthesis providers. Taking this step for government-funded research would be introducing a new requirement (where before there was guidance), but this step would promote screening as a norm, and there would be less comparative burden for gene synthesis providers that now do screen, as all scientists would need to use companies that screen in order to get grants.

In the United States, most researchers affected by such a requirement would be extramural researchers funded through NIH or the Department of Defense (DOD). They would be required to purchase gene synthesis products from companies that screen their orders, to include IGSC members and other companies that declare that they screen their orders. In 2018, NIH and DOD accounted for 27% and 39%, respectively, of federally awarded research and development funding.⁶⁸ Requiring federally funded researchers to acquire their DNA synthesis products from companies that perform screening would account for a significant number of life scientists who purchase DNA in the United States. These requirements would rapidly bring a large amount of research under customer and sequence screening, where it may not have been previously. This system could be as flexible or rigid as governments desire; certain agencies could be included or excluded from this process depending on different research or funding profiles.

Such a step may also have the effect of increasing the number of companies that perform screening. The IGSC maintains that their companies account for approximately 80% of the global gene synthesis capabilities. The remaining 20% that does not explicitly screen now would be incentivized to adopt screening practices to maintain their customer base. What will need to be worked out is how companies could demonstrate to federal funders of research that they screen their orders according to some minimum standard. Creating this kind of requirement for sourcing gene synthesis products from companies that screen would create a level playing field for providers and encourage screening.

Governments should require minimum standards for screening, but not be prescriptive about the specific database or sequence of concerns to be used for screening. Minimum standards should encompass regulated pathogens (such as on the Federal Select Agent Program lists and the Australia list). Recommending standards as compared to specific proprietary databases should encourage innovation and refinement of databases and approaches to screening. These standards should encompass already regulated pathogens. Going forward, standards might include additional sequences of concern that are identified by governments and their partners.

Some analysts have favored the development of a central database for screening, which all companies could use, simplifying the screening process. If there were a central database that was reliable, had proper annotations for sequences, and contained the pathogens of concern, a central database could theoretically reduce the time an analyst would need to investigate a hit. This would then reduce the overall costs, primarily by reducing the time needed for investigation. However, there are potential pragmatic problems with that approach that deserve further analysis. If there is just one database that all companies use, it may be more straightforward for a nefarious actor to evade screening, should the specific details of the database become known. As gene synthesis is a global industry, the regulated pathogens may vary from country to country as well. In addition, an attempt to have all gene synthesis orders be governed by a single database seems likely to be resisted or ignored by some scientists and some governments who were not involved in its creation or curation. During the Johns Hopkins Center for Health Security April 29 meeting, several experts remarked that implementing a shared-database approach would require a level of trust that is not now seen among several peer competitor nations.²⁹

In addition, keeping any database up to date would require continuous maintenance, annotation, verification, error correction, and funding, so a shared database would need to share in those activities as well. That work will be as complex and challenging as the original database creation. Overall, the work of database creation and management for synthesis screening efforts seems more suitable for the private sector than governments, given the pace of change and the need to have practicing scientists and informaticists involved in the work of maintaining and curating the database.

Governments should clarify that desktop synthesizer companies are considered “providers” of gene synthesis, with attending obligations. In the United States, the 2010 HHS Guidance states that “Providers should know to whom they are distributing a product. … Providers should know if the product that they are synthesizing and distributing contains, in part or in whole, a ‘sequence of concern.’” In addition, the 2010 HHS Guidance reminds providers of their obligations under federal law, including the responsibilities to adhere to the select agent regulations. To comply, gene synthesizers would require built-in biosecurity controls. For example, if a researcher wished to create a gene synthesis product that matched a virus on the Select Agents list, the researcher would encounter a non-skippable message on their synthesizer with instructions to contact the provider company for a clearance code to proceed. This step would require additional work and analysis to determine how biosecurity and functionality could be maintained if, for example, the synthesizer company should go out of business.

Governments should clarify that third-party companies that purchase genes and modify them for other applications and sell to specialized customers should be considered “providers” and adhere to the gene synthesis guidelines. Companies and institutions may outsource DNA synthesis or modification of synthesized products. For instance, if a DNA synthesizer is out of service, or the DNA synthesis core cannot synthesize the sequence as requested, they may outsource it to a third-party company.⁶² With the advent of DNA synthesis as digital storage space, companies may then outsource the synthesis of the designed sequence.⁶⁹ These third-party companies should also be considered providers so that they are held to guidelines for DNA synthesis and customer screening.

Governments should fund the development of screening methodologies and standards that could allow for the cost-effective screening of oligonucleotides. As technologies that rely on oligonucleotide synthesis to assemble larger pieces of DNA become more common, the need for screening lengths of DNA less than 200 nucleotides in length becomes more important. Funding the development of such screening would be a proactive effort for governments to maintain security in gene synthesis technology.

The US government, working with other governments, should find options to address the problems of customer screening in an international business context. For companies operating in the United States and serving US labs, customer information can be screened against US databases of proscribed persons or entities. But for synthesis screening in which the provider or the customer is located outside the United States, customer screening is more complicated. Does a synthesis provider located in country A, who is providing synthesis products for a customer in country B, use a proscribed person database from country A or country B? What if the provider or the customer is located in a country where there is no proscribed person database? To address this issue, the US government should convene other government representatives to develop common approaches to customer screening, perhaps in part by examining the practices used in other industries to screen customers.

The US government should actively engage with other countries, to encourage broader adoption of gene synthesis screening. The United States and other governments who are leading on these issues of synthesis screening should encourage the adoption of screening requirements internationally. The United States has learned a variety of practical lessons about what works around screening since the guidance was put in place in 2010. For example, it is valuable to have a government entity that companies can turn to if they have concerns that arise in the screening process. In the United States, the FBI WMD Directorate performs this function, and it would be useful for other countries that put screening requirements in place to have such a technical and investigative body to engage. These and other kinds of lessons could help other countries build their own effective screening programs. Few countries beyond the United States are yet paying close attention to gene synthesis screening in their guidance and regulations; the United States has an opportunity to promote the norm of screening, with broad benefits to biosecurity.

As a biosecurity tool, gene synthesis sequence and customer screening may be effective means of preventing misuse by some non-state actors, and despite rapid technological change, this norm should be maintained for as long as possible. Gene synthesis screening is not a biosecurity cure-all. It will not prevent sophisticated actors from acquiring pathogens they should not have access to and will not prevent actors from designing new versions of pathogens. However, protecting gene synthesis companies from having their services easily and directly misused will be a deterrent, and it will make acquiring dangerous pathogens more difficult.

References

Kaiser

Gobind Khorana and the rise of molecular biology. MIT School of Science. May 2, 2018. https://science.mit.edu/gobind-khorana-molecular-biology/. Accessed August 26, 2019.

Dorn

, Hall

, Hüsken

, et al. A role for oligonucleotide-based RNA-knock down technologies in functional genomics. Nucleosides Nucleotides Nucleic Acids, 2003; 22(5-8):641-644.

Hughes

, Ellington

. Synthetic DNA synthesis and assembly: putting the synthetic in synthetic biology. Cold Spring Harb Perspect Biol, 2017; 9(1):a023812.

Diggans

, Leproust

. Next steps for access to safe, secure DNA synthesis. Front Bioeng Biotechnol, 2019; 7:86.

Kosuri

, Church

. Large-scale de novo DNA synthesis: technologies and applications. Nat Methods, 2014; 11(5):499-507.

GENEWIZ. Gene Synthesis. https://www.genewiz.com/en/Public/Services/Gene-Synthesis. Accessed August 7, 2019.

Integrated DNA Technologies. Ultramer DNA Oligonucleotides. https://www.idtdna.com/pages/products/custom-dna-rna/dna-oligos/ultramer-dna-oligos. Accessed October 24, 2019.

Bergin

Synthetic Biology: Global Markets. BCC Research; July 2018. https://www.bccresearch.com/market-research/biotechnology/synthetic-biology-global-markets.html. Accessed November 1, 2019.

Wilson

Global gene synthesis market size, share, techniques, statistics, trends, new prospects, analysis, projections, insights and forecast to 2023. VC NewsNetwork February 6, 2019.https://www.reuters.com/brandfeatures/venture-capital/article?id=81330. Accessed August 26, 2019.

10.

Cello

, Paul

, Wimmer

. Chemical synthesis of poliovirus cDNA: generation of infectious virus in the absence of natural template. Science, 2002; 297(5583):1016-1018.

11.

Belser

, Tumpey

. The 1918 flu, 100 years later. Science, 2018; 359(6373):255.

12.

McMullan

, Flint

, Chakrabarti

, et al. Characterisation of infectious Ebola virus from the ongoing outbreak to guide response activities in the Democratic Republic of the Congo: a phylogenetic and in vitro analysis. Lancet Infect Dis, 2019; 19(9):1023-1032.

13.

Noyce

, Lederman

, Evans

. Construction of an infectious horsepox virus vaccine from chemically synthesized DNA fragments. PLoS One, 2018; 13(1):e0188453.

14.

Centers for Disease Control and Prevention. Federal Select Agent Program. Select Agents and Toxins List. https://www.selectagents.gov/SelectAgentsandToxinsList.html. Accessed August 7, 2019.

15.

National Center for Biotechnology Information. https://www.ncbi.nlm.nih.gov/. Accessed August 29, 2019.

16.

Matsui

, Corey

. Non-coding RNAs as drug targets. Nat Rev Drug Discov, 2017; 16(3):167-179.

17.

National Academies of

Sciences

, Engineering, and Medicine; Division on Earth and Life Studies; Board on Chemical Sciences and Technology; Board on Life Sciences; Committee on Strategies for Identifying and Addressing Potential Biodefense Vulnerabilities Posed by Synthetic

Biology

. Biodefense in the Age of Synthetic Biology. Washington, DC: National Academies Press; 2018.

18.

Gibson

DG.

Synthesis of DNA fragments in yeast by one-step assembly of overlapping oligonucleotides. Nucleic Acids Res, 2009; 37(20):6984-6990.

19.

Maurer

, Lucas

, Terrell

From understanding to action: community-based options for improving security and safety in synthetic biology. Berkeley: University of California, Berkeley; 2006. https://pdfs.semanticscholar.org/b488/e96eebaf688a811112ad9de890ea2c08a1a1.pdf?_ga=2.207514387.1226968914.1572622558-1372029453.1572622558. Accessed November 1, 2019.

20.

Fischer

, Maurer

. Harmonizing biosecurity oversight for gene synthesis. Nat Biotechnol, 2010; 28(1):20-22.

21.

Wimmer

The test-tube synthesis of a chemical called poliovirus. The simple synthesis of a virus has far-reaching societal implications. EMBO Rep, 2006; 7(Spec No):S3-S9.

22.

Randerson

Revealed: the lax laws that could allow assembly of deadly virus DNA. Guardian June 14, 2006. http://www.theguardian.com/world/2006/jun/14/terrorism.topstories3. Accessed October 30, 2019.

23.

Jones

Sequence screening. In: Garfinkel MS, Endy D, Epstein GL, Friedman RM, eds. Working Papers for Synthetic Genomics: Risks and Benefits for Science and Society. J. Craig Venter Institute, CSIS, MIT; 2005. https://www.jcvi.org/sites/default/files/assets/projects/synthetic-genomics-options-for-governance/Jones-Sequence-Screening.pdf. Accessed November 1, 2019.

24.

Garfinkel

, Endy

, Epstein

, Friedman

. Synthetic Genomics Options for Governance. J. Craig Venter Institute; 2007. https://www.jcvi.org/synthetic-genomics-options-governance. Accessed November 1, 2019.

25.

US Department of Health and Human Services. Screening Framework Guidance for Providers of Synthetic Double-Stranded DNA. Last reviewed May 4, 2015. https://www.phe.gov/Preparedness/legal/guidance/syndna/Pages/default.aspx. Accessed June 24, 2019.

26.

National Center for Biotechnology Information. GenBank Overview. https://www.ncbi.nlm.nih.gov/genbank/. Accessed September 3, 2019.

27.

The Australia Group. https://australiagroup.net/en/. Accessed September 3, 2019.

28.

International Gene Synthesis Consortium. https://genesynthesisconsortium.org/. Accessed June 24, 2019.

29.

Johns Hopkins Center for Health Security. Policy options for gene synthesis. Johns Hopkins Center for Health Security. 2019. http://www.centerforhealthsecurity.org/our-work/Center-projects/policy-opts-gene-synthesis.html. Accessed October 30, 2019.

30.

Gibson

, Young

, Chuang

, Venter

, Hutchison

CA 3rd

, Smith

. Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods, 2009; 6(5):343-345.

31.

Deng

, Chen

, Liao

, et al. A simplified and versatile system for the simultaneous expression of multiple siRNAs in mammalian cells using Gibson DNA assembly. PLoS One, 2014; 9(11):e0113064.

32.

Czar

, Anderson

, Bader

, Peccoud

. Gene synthesis demystified. Trends Biotechnol, 2009; 27(2):63-72.

33.

, Elledge

. Harnessing homologous recombination in vitro to generate recombinant DNA via SLIC. Nat Methods, 2007; 4(3):251-256.

34.

Quan

, Tian

. Circular polymerase extension cloning of complex gene libraries and pathways. PLoS One, 2009; 4(7):e6441.

35.

Zhang

, Werling

, Edelmann

. SLiCE: a novel bacterial cell extract-based DNA cloning method. Nucleic Acids Res, 2012; 40(8):e55.

36.

Gibson

, Benders

, Axelrod

, et al. One-step assembly in yeast of 25 overlapping DNA fragments to form a complete synthetic Mycoplasma genitalium genome. Proc Natl Acad Sci U S A, 2008; 105(51):20404-20409.

37.

Palluk

, Arlow

, de Rond

, et al. De novo DNA synthesis using polymerase-nucleotide conjugates. Nat Biotechnol, 2018; 36(7):645-650.

38.

Molecular Assemblies expands patent portfolio in enzymatic DNA synthesis [news release]. Molecular Assemblies September 18, 2018. https://molecularassemblies.com/molecular-assemblies-expands-patent-portfolio-enzymatic-dna-synthesis/. Accessed August 12, 2019.

39.

Feldman

, Romesberg

. Expansion of the genetic alphabet: a chemist's approach to synthetic biology. Acc Chem Res, 2018; 51(2):394-403.

40.

Piccirilli

, Krauch

, Moroney

, Benner

. Enzymatic incorporation of a new base pair into DNA and RNA extends the genetic alphabet. Nature, 1990; 343(6253):33-37.

41.

Malyshev

, Dhami

, Lavergne

, et al. A semi-synthetic organism with an expanded genetic alphabet. Nature, 2014; 509(7500):385-388.

42.

Hoshika

, Leal

, Kim

, et al. Hachimoji DNA and RNA: a genetic system with eight building blocks. Science, 2019; 363(6429):884-887.

43.

Warren

Four new DNA letters double life's alphabet. Nature, 2019; 566(7745):436.

44.

Feldman

, Dien

, Karadeema

, et al. Optimization of replication, transcription, and translation in a semi-synthetic organism. J Am Chem Soc, 2019; 141(27):10644-10653.

45.

Fredens

, Wang

, de la Torre

, et al. Total synthesis of Escherichia coli with a recoded genome. Nature, 2019; 569(7757):514-518.

46.

Gros

, Martínez-Quintanilla

, Puig

, et al. Bioselection of a gain of function mutation that enhances adenovirus 5 release and improves its antitumoral potency. Cancer Res, 2008; 68(21):8928-8937.

47.

BioAutomation. MerMade 192E. https://bioautomation.com/products/192e. Accessed October 24, 2019.

48.

SGI-DNA. BioXp^TM 3200 System. https://www.sgidna.com/products/bioxp/. Accessed August 7, 2019.

49.

Hyde

The future of DNA synthesis. SynBioBeta January 1, 2019. https://synbiobeta.com/the-future-of-dna-synthesis/. Accessed August 7, 2019.

50.

Carlson

The changing economics of DNA synthesis. Nat Biotechnol, 2009; 27(12):1091-1094.

51.

Gene Universal. Custom gene synthesis. http://www.geneuniversal.com/. Accessed October 24, 2019.

52.

GenScript. Gene synthesis & DNA synthesis services. https://www.genscript.com/gene_synthesis.html. Accessed October 24, 2019.

53.

John

Gene synthesis service market 2019 consumption forecast & business region distribution analysis to 2026 by Payers Genescript, GeneArt (Thermofischer), IDT, Eurofins Genomics, DNA 2.0 (ATUM), BBI, Genewiz, Gene Oracle. Market Report Gazette September 22, 2019. https://www.marketreportgazette.com/2019/09/gene-synthesis-service-market-2019-consumption-forecast-business-region-distribution-analysis-to-2026-by-payers-genescript-geneart-thermofischer-idt-eurofins-genomics-dna-2-0-atum-bbi/. Accessed October 30, 2019.

54.

Global DNA Read, Write and Edit Market. Research and Markets. September 2019. https://www.researchandmarkets.com/reports/4833948/global-dna-read-write-and-edit-market. Accessed October 24, 2019.

55.

Carter

, Friedman

. DNA Synthesis and Biosecurity: Lessons Learned and Options for the Future. La Jolla, CA: J. Craig Venter Institute; 2015.

56.

Battelle. ThreatSEQ^TM DNA Screening Web Service. https://www.battelle.org/commercial-offerings/industry-solutions/threatseq-dna-screening-web-service. Accessed August 12, 2019.

57.

Twist Bioscience. Twist Bioscience adopts Battelle's ThreatSEQ^TM DNA Screening Web Service for Advanced Biosecurity [news release]. Twist Bioscience April 9, 2019. https://investors.twistbioscience.com/news-releases/news-release-details/twist-bioscience-adopts-battelles-threatseqtm-dna-screening-web. Accessed August 12, 2019.

58.

IARPA. Functional Genomic and Computational Assessment of Threats (Fun GCAT). https://www.iarpa.gov/index.php/research-programs/fun-gcat. Accessed October 30, 2019.

59.

IARPA. Finding Engineering-Linked Indicators (FELIX). https://www.iarpa.gov/index.php?option=com_content&view=article&id=999&Itemid=397. Accessed October 30, 2019.

60.

DARPA. Safe Genes. https://www.darpa.mil/program/safe-genes. Accessed October 30, 2019.

61.

Ranomics. Powering discoveries using genomics. https://ad.ranomics.com/genesynthesis/?gclid=EAIaIQobChMIq7K4m5jv4wIVj56fCh3GEAPhEAMYASAAEgLZJPD_BwE. Accessed August 12, 2019.

62.

Massachusetts General Hospital. MGH CCIB DNA Core. DNA Synthesis: Custom Oligonucleotides. https://dnacore.mgh.harvard.edu/new-cgi-bin/site/pages/synthesis_pages/syn_custom_oligos.jsp. Accessed August 26, 2019.

63.

Massachusetts General Hospital. MGH CCIB DNA Core. DNA Synthesis. https://dnacore.mgh.harvard.edu/new-cgi-bin/site/pages/synthesis_main.jsp. Accessed August 12, 2019.

64.

Massachusetts General Hospital. MGH CCIB DNA Core. DNA Synthesis: Oligonucleotide Purification. https://dnacore.mgh.harvard.edu/new-cgi-bin/site/pages/synthesis_pages/syn_oligo_purification.jsp. Accessed August 26, 2019.

65.

Ranomics. Genomics and screening solutions. 2019. https://www.ranomics.com. Accessed August 26, 2019.

66.

Gibson

, Glass

, Lartigue

, et al. Creation of a bacterial cell controlled by a chemically synthesized genome. Science, 2010; 329(5987):52-56.

67.

DiEuliis

, Carter

, Gronvall

. Options for synthetic DNA order screening, revisited. mSphere, 2017; 2(4):e00319-17.

68.

Congressional Research Service. Federal Research and Development (R&D) Funding: FY2020. Washington, DC: Congressional Research Service; 2019. https://fas.org/sgp/crs/misc/R45715.pdf. Accessed November 1, 2019.

69.

Hysolli

Save it in DNA. Wyss Institute. August 6, 2019. https://wyss.harvard.edu/news/save-it-in-dna/. Accessed October 24, 2019.