comment 0

Kill switches prevent escape of genetically modified bacteria

bacterial kill switch

There is considerable public concern about the use of genetic engineering, particularly around the production of food. The most recent example is the FDA-approved, fast growing salmon from the company AquaBounty. This fast growing fish is grown in land-locked tanks as sterile females so that they cannot reproduce if they were to escape into the wild. Interestingly, this fish grows almost twice as fast as salmon in the wild due to the increased production of a fish growth hormone, genetically engineered to be expressed from multiple organs instead of in limiting amounts from the pituitary gland. The basic research that led to this transgenic fish was described over 20 years ago from a Canadian researcher, Dr Garth Fletcher of Memorial University.

What is less appreciated by the public are the numerous examples of genetic modification that result in medicines or other useful and less controversial products. One of the first products resulting from genetic modification was the production of synthetic insulin used to treat diabetes, which is expressed and purified from bacteria.  And the exploding field of synthetic biology has the potential of providing a bounty of new products, medicines, bioremediation strategies using engineered bacteria or biofuel production by microbes. If genetically engineered bacteria are to be employed in the field, at sites of contamination, or in production factories, there is a significant worry about their escape and the possible implications of release into the wild. To address this concern, numerous research groups are designing so-called ‘kill switch’ strategies, which will prevent the growth of these microbes in the environment. A creative new strategy was described earlier this year.

Microbiologists frequently work with ‘auxotrophs’, which are strains that cannot synthesize a specific nutrient, and therefore do not grow without its supply to the microbe. As standard auxotrophs might be capable somehow of scavenging these needed nutrients, ‘synthetic auxotrophs’ were designed that rely on the supply of unusual amino acids that are not commonly found in nature. There are several uncommon amino acids that are coded for in the triplet DNA code, but that are not found very often in proteins. This suggests the possibility to change the genetic code and introduce new codons that would introduce  non-standard amino acids into proteins.

The first stage of the kill switch technology was founded on an earlier landmark paper that used whole genome editing,  where all 321 UAG stop codons were replaced with UAA codons (ref 1). UAG was convenient to alter, since it appears rarely in the genome and only at the end of genes. These strains are referred to as genetically recoded organisms, or GROs.  Note, there were 355 off target mutations created in the process, which did not affect viability but may need to be corrected in future. It should be noted that this GRO strain was now quite resistant to infection from viruses, another advantage to this type of biocontainment strategy. The new genetic code mistranslates viral proteins that hijack the bacterial protein synthesis machinery, which ultimately disables new virus assembly leading to resistance.

Several other sophisticated, genetic manipulations were performed in order to create a growth dependence. It was now possible to reassign UAG as a dedicated sense codon for “plug and play” incorporation of rare amino acids. In order to translate the UAG codon to a  meaningful response by the protein synthesis machinery,  they expressed a specific tRNA synthetase that recognizes the UAG codon and introduces  1-4,4′-biphenylalanine (bipA) into target proteins during synthesis. As there were no longer any functional UAG codons in the genome, they selected a handful of essential proteins and redesigned these genes to encode UAG at various positions. Structure predictions were used to select the candidate essential genes to recode, which were confirmed by protein structure determinations. Mass spectrometry also confirmed the introduction of bipA into target proteins.

 Unfortunately, some synthetic auxotroph strains were still viable at low frequencies when grown in the the absence of bipA, indicating that compensatory mutations occurred which restored viability. In other words, the supposed ‘kill switch’  was not perfect in the earliest attempts. Several additional mutations were needed, including the addition of conventional auxotrophs, to really limit growth and escape of this strain in the absence of the non-standard amino acids. They ultimately identified a highly engineered strain could only survive and produce functional essential proteins when grown in the presence of the non-standard amino acid, 1-4,4′-biphenylalanine (ref 2). This ‘tour de force’ approach created a synthetic auxotroph which now can now be used as the ‘chassis’ for the future development of a particular biotechnology.

In the same issue, a second paper on the same subject was also published. They generally used a similar approach to construct synthetic auxotrophs that were dependent on the exogenous supply of unnatural analogs of phenylalanine (ref 3). The end result was the same, a strain that could not escape its amino acid dependency through mutation or the acquisition of multiple genes through horizontal gene transfer. The NIH standard for an effective kill switch is an approach where survivors do not occur with a frequency of 1 per 108 cells. Other non-auxotroph strategies have been used that meet this minimum requirement.  Environmental microbes capable of degrading environmental pollutants have been designed to express a lethal toxin when in the environment, but only once the pollutants have been degraded. This permits the deliberate use of bacteria for bioremediation of contaminated soils but limits their survival after the job is done. These bacterial suicide strategies were developed to mitigate possible risks and should permit the safe use of microbes in the environment to help resolve environmental or health crises, with the benefit greatly outweighing the risk.

Ref 1. Genomically Recoded Organisms Expand Biological Functions. 10.1126/science.1241459

Ref 2. Biocontainment of genetically modified organisms by synthetic protein design. doi:10.1038/nature14121

Ref 3.  Recoded organisms engineered to depend on synthetic amino acids. doi:10.1038/nature14095


comment 0

Cloud-based lab experiments

twist bioscience dna synthesis

Image source.

The field of synthetic biology is seeing major interest in the areas of DNA synthesis and automation. I’m not talking about liquid handling stations, but complex automated workstations that assemble genetic constructs, transform microbes, perform high throughput experiments, automated analysis, all enabled with sophisticated visualization tools. The vision is that experiments can be first designed, performed by workstations that are controlled by computers through the cloud, and analyzed, all within the span of a few days. Everything can be done from your computer, which brings great promise of eliminating the grunt work typically associated with molecular biology. We still hear the old rumours of colleagues who used to prepare their own enzymes for PCR. Pretty soon, we will be laughing at those who still show up to the lab to operate PCR machines, running DNA and protein gels.

Several interesting companies are seeking to occupy this space. Emerald Cloud Labs is one company that offers the whole package. The list of possible experiments that can actually be performed is quite impressive, and includes DNA/RNA synthesis, microscopy, cell culture, and analytical methods. The long list of experimental options can be seen here, and the company vision can be seen here. One of first companies on the automated platform biofoundry scene is Gingko Bioworks.  They recently partnered with DNA synthesis company Twist Biosciences, in a landmark deal where Gingko purchased 100 million base pairs of synthetic DNA, which is equivalent to 10% of the current DNA synthesis capacity worldwide. This combined ability to synthesize DNA fast and cheap, coupled to the bioworks automation, should result in the creation of an impressive number of new products. In their pipeline of partners, they are working on engineered organisms to produce sweeteners, fragrances, cosmetics and flavors. They also expect to contribute to designing microbes that consume CO2 and convert it to biofuels, or produce new probiotics, and also to identify new medicinal compounds from screening high throughput drug libraries and natural compounds.

Personally, I find these advances awe inspiring. I’ve always been attracted to applied microbiology research. These projects will require greater efforts in the design stage, but the pace of discovery will come fast and furious. As someone now working for a distance education university, the ability to contract out complicated, time consuming experiments is a breath of fresh air. I think we are going to see a new structure to budget sections of grants, where much of the work is a single line item contracted out to a third party. Welcome to the future!

comment 0

Brewing morphine from glucose in yeast

poppy field

The opium poppy plant Papaver somniferum is currently used as the sole commercial source of the most powerful therapeutic, prescribed opiate pain killers. There has been a long time interest in using synthetic biology to produce opiates as bioproducts, with the goal of controlling the drug, and hopefully producing cheaper, less addictive and more secure medicines. The family of molecules known as benzylisoquinolone alkaloids (BIA) include many plant products derived from tyrosine, such  as pain killers, antimicrobials, anti-cancer drugs  muscle relaxants and cough suppressants.

Within the first six months of 2015, three reports have described a complete biosynthetic pathway using yeast to convert glucose into morphine1,2,3.  Although there have been numerous previous advances in the last several years, it is now possible to engineer the entire morphine biosynthesis pathway in yeast in order to brew morphine instead of beer. While most enzymes in the pathway were derived from the poppy plant, there was also the use of other plant and bacterial enzymes. The ability to combine these enzymes in a single engineered strain still requires significantly more work, but it does bring to the forefront numerous ethical issues. The most intriguing issue revolves around how to prevent these engineered strains from being obtained or possibly reproduced by the drug cartels. Although this area of research has generated gossip of another season of the television series Breaking Bad, featuring a chemistry teacher turned methamphetamine cook and dealer, we will focus here on the synthetic biology behind these discoveries.

Glucose to (S)-Reticuline

During normal metabolism and growth of yeast on glucose, the amino acid tyrosine is naturally produced during the process of amino acid synthesis. The researchers first tackled a long standing problem in yeast, converting tyrosine to L-DOPA1. A plant enzyme was introduced that converts L-DOPA to a fluorescent pigment, whose production can be easily monitored in a biosensor assay looking for pigmented colonies, or using fluorescence to detect yellow yeast cells. It was initially expected that a high throughput screen of cDNA libraries would be necessary, but a bioinformatic analysis revealed two candidate enzymes, one from the button mushroom and another from the sugar beet, that might convert tyrosine to L-DOPA. Codon optimized versions of these enzymes were introduced and yeast strains expressing the sugar beet enzyme were shown to produce L-DOPA, which was converted to the yellow pigment. Further enzyme engineering was carried out using error-prone PCR to generate a library of mutant enzymes, which were successfully screened for variants that had improved yields of L-DOPA1.

The next step was to produce (S)-Reticuline, a central precursor in the biosynthesis of opiates. Using a combination of one bacterial and several poppy enzymes, most of which were previously identified and many shown to function in yeast, (S)-Reticuline was produced. This required the introduction of many non-yeast genes, expressed from strong promoters either from the chromosome or low-copy plasmids, and the use of unique yeast background strains that produced higher than normal amounts of key metabolites required for (S)-reticuline synthesis.

(R)-Reticuline to Morphine

Using knowledge of the morphine biosynthesis pathways in the poppy, a yeast strain was engineered to express 7 poppy plant enzymes, which led to the production of the critical intermediate thebaine, and ultimately the final products, codeine and morphine2.  Cultures were supplemented with (R)-reticuline, because at this time the enzyme that converted (S) to (R)-reticuline had not been described. This kind of pathway engineering can also be tweaked to produce other important final products derived from thebaine, including  opiod antagonists, which are used to treat opiate addiction. Similarly, thebaine is also a starting point for the chemical synthesis of oxycodone, which has a better side-effects profile, compared to morphine.

The most recent paper identified the poppy enzyme responsible for the S-to-R epimerization of reticuline3. With this finding, all enzymes required for morphine biosynthesis were identified. The pathway contains  fifteen enzymatic steps, compared to a previous synthetic biology accomplishment of expressing five enzymes needed for the synthetic production of the antimalarial compound, artemisin.

Final thoughts on brewing morphine.

While the current morphine production from the poppy plant does meet demand, synthetic biologists have made great progress in synthesizing this compound in both bacteria in yeast.  In comparison, bacteria still make higher yields of important intermediates like L-DOPA, when compared to yeasts,  but the overall  yield of opiates are still very low. However, the stage is set to produce a single yeast strain with all necessary enzymes and to optimize the production of high morphine yields from yeast fermentation reactions.

Another major advance in the field of synthetic biology is the ability to synthesize custom stretches of DNA. The relative low cost to order “DNA to go” may lead to abuses from individuals with sinister motives in attempting to create bioweapons, and possibly to attempt to construct their own morphine brewing microbial strains. DNA synthesis companies do regular screening of the sequences they produce, and it has been suggested to now search these orders for enzymes in the morphine synthesis pathway.


1.  An enzyme-coupled biosensor enables (S)-reticuline production in yeast from glucose. DeLoache WC, Russ ZN, Narcross L, Gonzales AM, Martin VJ, Dueber JE. Nat Chem Biol. 2015 Jul;11(7):465-71. doi: 10.1038/nchembio.1816. Epub 2015 May 18.

2. Synthesis of Morphinan Alkaloids in Saccharomyces cerevisiae. Fossati E, Narcross L, Ekins A, Falgueyret JP, Martin VJ. PLoS One. 2015 Apr 23;10(4):e0124459. doi: 10.1371/journal.pone.0124459.

3. Stereochemical inversion of (S)-reticuline by a cytochrome P450 fusion in opium poppy. Farrow SC, Hagel JM, Beaudoin GA, Burns DC, Facchini PJ. Nat Chem Biol. 2015 Jul 1. doi: 10.1038/nchembio.1879

Canadian Synthetic Biology Researchers

Vincent Martin, Professor, Concordia University

Peter Facchini, Professor, University of Calgary