Although synthetic biology holds the potential to revolutionize everything from healthcare to sustainable manufacturing, it has also spawned fears of engineered organisms escaping the lab and wreaking havoc. Ever since Viktor Frankenstein first plugged life into his unwitting subject, something about manipulating life has held a special place in humanity’s collective nightmares.
Early geneticists, recognizing this, called a conference in Asilomar, CA in 1975, where they agreed on a set of guidelines for the spanking new technology they called “recombinant DNA”. Since then, the NIH has issued formal guidelines on biocontainment of genetically modified organisms (GMOs).
Among the definitions and appendices and exceptions and exemptions stands one golden number: 10-8. This is the threshold set for the escape probability of potentially dangerous genes; if 108 engineered cells escaped the lab, only 1 would either survive or pass on its modified genes.
The quest for 10-8: Kill switches and auxotrophic containment
Most standard lab experiments don’t really pose any risk and are exempt from these guidelines. For the rest, biologists have developed several strategies to try and meet the 10-8 escape threshold. One strategy is a built-in “kill switch” – the GMO could be given a toxin/antitoxin gene pair, so it would constantly be both poisoning itself with a toxin gene and saving itself with the antidote. If the antitoxin gene was only turned on when the GMO was given a specific inducing chemical, then outside the lab, the antidote would turn off, activating the GMO self-destruct sequence. However, this system is relatively easy to escape by mutating out the toxin gene.
Another strategy is make the organism an auxotroph by knocking out the GMO’s ability to make some life-essential molecule, making it, like grad students, dependent on some source of free food. In the lab, this auxotrophic GMO would get its handout from lab media. Unfortunately, this strategy doesn’t work very well either. Again like grad students, auxotrophs can often find free food sources outside the lab too, such as nutrients leaking from other cells or cookies at a department seminar.
In addition to GMOs escaping kill-switch or auxotrophic containment, the modified genes themselves could potentially escape into another organism. The potential for escape of engineered organisms and genes has prompted significant thought and work on how to restrain them.1,2,3,4
In a pair of new studies, the labs of George Church and Farren Isaacs demonstrated an impressive new way to get under the 10-8 threshold – they made GMOs that use a new genetic code.
A new solution: genetically recoded organisms
The basic idea is a variant of the auxotrophy approach, except that instead of making the cells depend on a nutrient they could find in nature, they would need a completely synthetic amino acid, one not made by any living thing outside of a white coat. In theory, that should prevent the live cells from escaping, and if the potentially dangerous genes also needed this synthetic amino acid, the genes wouldn’t be able to escape either.
To make this “genetically recoded organism” (GRO), the two labs started with a domesticated lab strain of E. coli that had previously had all of its TAG stop codons replaced with TAA, along with removal of the release factor that stops mRNA translation at TAG. That freed up the TAG codon for use with a completely artificial amino acid. George Church’s lab used the hydrophobic amino acid bipA, which they inserted into several essential enzymes, along with predicted secondary mutations to compensate for its different shape.
Meanwhile, Farren Isaacs’ lab used another synthetic amino acid, pAcF, which they inserted at various positions where they predicted it wouldn’t break the enzyme. Both labs successfully created GROs with new amino acid letters in their genetic vocabularies.
Genetically Recoded Organisms Aren’t Perfect
However, the work was not yet done. Even when the cells’ essential enzymes required completely synthetic amino acids, some still escaped when switched to media without those synthetic nutrients. Some cells escaped by mutating their new TAGs to another codon that uses a standard amino acid. Others escaped through “amber suppressor” mutations, which change a standard tRNA to recognize TAG. Still others escaped through various other random mutations.
Also, when the GROs were allowed conjugal visits with wild-type bacteria that were competent to transfer DNA, the visitors helped their partners escape by sneaking them wild-type versions of the altered genes.
There was still hope, though, for the biocontainment jailers. By altering several genes in the same strain, both groups achieved escape frequencies that were not only below the magical 10-8 , but were also below their limits of detection: 10-11 to 10-12. Escape via conjugal transfer of DNA might not be completely preventable, but if any potentially dangerous genes were sandwiched in between the containment genes, they would be removed along with the containment.
<10-8: The future for biocontainment
In the end, the two groups dramatically improved biocontainment over current strategies. Their approach does still run into the “life will find a way” principle enshrined in another of pop culture’s biological nightmares (albeit a beautifully scored one), Jurassic Park.
However, the labs aren’t likely to stop with just one synthetic amino acid per strain. Right now, their GROs don’t so much speak a different language from natural life as just a different dialect. By combining multiple layers of altered enzymes and synthetic amino acids, the decimal point in the escape frequency might move far enough to the left that we can just call it zero.
While this is impressive work, synthetic biologists aren’t likely to switch over to shackled GROs overnight. For one thing, the further their genetic language gets from natural life, the less we learn about natural life by studying them. Secondly, fears of GMO escape are probably overblown, since natural organisms are already exquisitely evolved, and almost all tinkering we do handicaps them out in the wild. However, this could eventually be quite useful for specific applications, like engineering strains that we want to stay outside but localized (e.g., for bioremediation or biofuel production), or for studying particularly bad diseases.
Whatever the eventual applications, we can recode life to speak another genetic language, and that’s pretty friggin cool.
Learn to read a new genetic code over in Mandell et al., Nature, January 2015 and Rovner et al., Nature, January 2015.
- Schmidt, M., de Lorenzo V. Synthetic constructs in/for the environment: Managing the interplay between natural and engineered Biology. FEBS Lett. 2012:586;2199–2206.
- Wright, O., Stan, G-B., Ellis, T. Building-in biosafety for synthetic biology. Microbiology 2013:159;1221–1235
- Moe-Behrens, G.H.G., Davis, R., Karmella, A. Haynes Preparing synthetic biology for the world. Front Microbiol. 2013:4:5.
- de Lorenzo V. Environmental biosafety in the age of Synthetic Biology: Do we really need a radical new approach? Bioessays 2010:32;926–931.