HOW SAFE IS “SAFE ENOUGH”?
1. IntroductionTo understand and to form an opinion about the safety of synthetic biology, people should know what synthetic biology is. These days, no global definition of this term exists: people all over the world give a slightly different interpretation. Our view is well displayed by the definition posed by the ETC group (a Canadian Action Group on Erosion, Technology and Concentration). For a good understanding of our safety analysis we give their view on this complex matter. Synthetic biology is ‘the design and construction of new biological parts, devices and systems that do not exist in the natural world and also the redesign of existing biological systems to perform specific tasks’ (the ETC group). For the creation of these synthetically engineered organisms we make use of biological building blocks, called BioBricks. While creating such organisms, scientists reflect on the possible consequences of their future creations. Even with the best intentions, experiments can go wrong or people with bad intentions can abuse the results.
While engineering E.D. Frosti we added several mechanisms to ensure biosafety. When a stimulus (arabinose or lactose) is combined with a cold temperature, a cell death mechanism will kill off the cells. Hereby the bacteria will be killed but the produced proteins stay active on their cell surface for some period of time. We thought about the dangers that E.D. Frosti could entail for the environment and even for mankind, if something would go wrong. But the question remains: How safe is “safe enough”?
2. Safety in the labThe Irish novelist S. Lover said:” It’s better to be safe than sorry”. While working in the lab, this is the key sentence that the K.U. Leuven iGEM team 2011 kept in mind. We are working in a laboratory in a laminar flow with biosafety level 1 because the host organisms we use are non-infectious. Every student working in the laboratory attended and passed the course ‘safety and laboratory practice'. We apply the standard rules of ‘Good Laboratory Practice’; we always wear a lab coat, safety goggles and if using chemicals or handling organisms, we wear gloves. In addition, a member of the safety institution of our university ‘Health-Safety-Environment’ ( https://admin.kuleuven.be/vgm/en ) is standby and the advisors are constantly present to guide, assist and help us. We reported the project to our department and HSE by filling out the forms ‘risk assessment for experiment with hazardous biological materials'. We asked the HSE to give feedback on the safety of our project, click here to read it. In this form we considered the danger of working with genetically modified organisms. In Belgium, we have the Belgian Biosafety Server who has formed some federal regulations and laws concerning lab safety. Besides those safety measures, the K.U.L. has its own guidelines and procedures that are strictly mandatory. The HSE makes sure that all this requirements are fullfilled.
3. Safety and the environmentMarcus Schmidt wrote a chapter in his book ‘The technoscience and its societal consequences’: “Do I understand what I can create?”  Synthetic biology opened a whole new world for scientists where they can create everything they have ever dreamed of. There are many remarkable ideas, like an organism that detects when people become ill and cures them immediately, such as Dr. coli, to bacteria that fill micro cracks in concrete such as the Bacillafilla. [3, 4] The aim of these projects was to help people by improving the quality of their life.
But we have to keep in mind that there are also people who may not have the best intentions when creating a genetically manipulated organism. Although biosecurity – the prevention of intentional release of pathogens and toxins – is not something we can control, it is important to think about what happens if our creation falls in the hands of the wrong people.  And finally, we consider that even when we have good intentions, we can still, unintentionally, create something very dangerous.
The strain of E. coli that we are using, MG1655, is a derivate of K-12 and has the same properties. E.coli K-12 normally doesn’t colonize the human intestine. This strain is commercially used and has not any known adverse effects on humans, animals or the environment. If these bacteria are exposed to the environment, they would have low survival chance. Indeed, they would have to compete with other organisms for nutritional sources and E. coli could lose this competition since the other organisms are adapted to their habitat, while ''E. coli'' is adapted to live in the mammalian digestive system. [5, 6, 7]
The basic idea of our project is to make the bacteria E.D. Frosti, which makes ice nucleating proteins or anti-freeze proteins, after which it kills itself. The bacteria themselves only act as physical carriers of the proteins. Afterwards, we would add the bacteria with the ice nucleating proteins on their membrane to cold water, which allows the water to freeze faster, and crystallize at a higher temperature. On the other hand, bacteria coated with the anti-freeze protein can be used to induce ice melting, so that they can replace sprinkling salt on the roads. One problem is if the general public will accept engineered bacteria in the environment. If something went wrong in our project, this could have major consequences for nature and mankind. We worked out our biosafety issues with the ‘Fault Tree Analysis theory’. This method defines unwanted scenarios and traces them back to the necessary precautions that are needed to avoid these failures.  We’ll demonstrate this theory with some examples relevant to our project.
Apocalyptic scenario 1: creation of the everlasting Ice age.
One of the possible scenarios is an everlasting ice age. This could happen if the bacteria would make nonstop ice nucleating proteins and find their way into canals, hereby freezing all connected canals. Frozen canals, rivers and lakes can lead to frozen oceans. We would need new technologies to defrost water so we can drink, shower, etc. But it doesn’t stop there: the ice cap will reflect more sunbeams and the temperature on earth will get lower. So eventually, the whole planet would turn to ice. This scenario would speed up the ‘snow ball earth hypothesis’.  Luckily, E.coli (and thus our E.D.Frosti) cannot survive at freezing temperatures, so cell death will occur.
Apocalyptic scenario 2: the final meltdown.
Instead of the ice nucleating protein staying active, imagine that the anti-freeze protein stays active and the bacteria reach the arctic. If an enormous amount of anti-freeze protein reaches this point, it could cause a glacial meltdown. This scenario is comparable to an acceleration of the global warming, leading to more floods, tsunamis, tornado, etc. and this could lead to the extinction of human race. 
Luckily, we have taken several measures to make sure that the cells cannot overgrow the environment, as stated above. We engineered E.D. Frosti to induce cell death slowly when encountering cold temperatures. Even when the cells are dead the proteins remain attached to the cell membrane and carry on their function during an acceptable amount of time. After some time, the proteins will degrade and the effect of our engineered bacteria will stop. This safety mechanism ensures that the apocalyptic scenarios above won’t happen. After the first stimulus of arabinose or lactose (transcription and translation of the ice nucleating protein or the anti-freeze protein), the DNase activity is induced when the cells encounter low temperatures. The DNase activity degrades the DNA and the cell dies. Thus, as mentioned before, the cells only function as a physical carrier for the proteins and most of the cells will die when exposed to a cold environment.
Deep ocean water has average temperatures between 0-3°C.  At these temperatures ice nuclear proteins are able to crystalize water.  However E. coli is unable to survive or multiply at low temperature and moreover the cell death mechanism of our E.D. Frosti will be induced. The same goes for E.D. Frosti expressing the anti-freeze protein that could reach the cold arctic. With this information we conclude that the probability of both the "everlasting ice age" and the “final meltdown” scenario, happening are small. The amount of bacteria needed for either scenario are that huge that the only way this could happen is if it were to be done on purpose. This is a biosecurity problem.
4. Safety and the publicApocalyptic scenario 3: Freezing from the inside out.
During the winter, after a heavy snow storm, salt is scattered to clear the roads. This salt induces corrosion of car parts and when it floods into the surrounding natural environment, the change in salinity of the soils can greatly impact the local fauna and flora. Instead of using salt, our bacteria E.D.Frosti could be used. This is environmentally friendlier than salt. A disadvantage however is that bacteria with their proteins remain in the environment for some period of time and could contaminate the water. If we ingest the bacteria and the ice nucleating protein is still active, ice crystallization in our body could cause organ damage.
If we assume the E.D. Frosti spreads throughout someone’s body and if that person drinks a lot of cold water, there is a high probability that the bacteria would instantly freeze the cold water. This could reduce the overall body temperature and eventually freeze the whole human body.
The possibility of this scenario happening is insignificant. When we swallow bacteria, they are able to survive until they get into the abdomen. In the stomach the pH lies between 1 and 2, causing the proteins to denaturize and break down into peptides.  If the bacteria don’t enter the stomach, the INP will stay active and will try to induce ice crystal formation. The protein will be active but won't cause any ice crystallization because the body temperature is too high to induce this.
Toxicity- “Prevention is better than cure”
During the initial design of our project we planned to use copper ion as the inducer of the INP protein pathway. Although this system would work very well as a proof of concept, we quickly realized that eventually the copper would be found in high toxic levels in the cell debris affecting people, animals, and the environment.
Copper toxicity leads to vomiting, low blood pressure, coma, jaundice, vomiting of blood, melena and gastrointestinal distress. Living organisms like mammals have well defined and optimized processes to regulate the excess of copper within certain ranges. However the body becomes poisoned when faced with too high concentrations. There are general safety standards concerning metal pollution in Europe and U.S. which enforce strict regulations on the amount of copper in fumes, particles and aerosols. While identifying these risks we decided to use sugar based inducers, such as lactose and L-arabinose, to trigger the INP pathway. These do not pose any toxicity risks.
The ice nucleating protein occurs naturally in the bacterium Pseudomonas syringae. Scientists did research on the toxicity of this bacterium and proved that this bacterium is not a pathogen for humans or animals.  Yet we decided not to use this organism because it is proven to be a plant pathogen.  Pseudomonas syringae is adapted to cold temperatures, we believe if this strain was able to survive the high pressure at the surface layer of the ocean it would be able to freeze a part of it.
5. Danger of exchange of DNAApocalyptic Scenario 4: The 'X-men revolution'.
E. coli naturally occurs in our intestinal flora. The optimal temperature for growth is 37°C and pH is 7-8. E.D. Frosti is genetically engineered starting from E. coli; therefore we assume the bacteria will have the same characteristics. If we were to swallow the bacteria and we would take in the foreign DNA, this could cause the ‘X-men revolution’. The “powers” someone would gain after this, are difficult to foresee. It could result in people who can control ice formation and defrosting by self-secreting the ice nucleating protein and the anti-freeze protein.
For this to take place the DNA would have to get through the nucleus and the cell membrane and find a way to escape or survive the acidity (pH 1-2) of the stomach. Given that the inside of the cell membrane is hydrophobic and DNA is a big hydrophilic molecule, this scenario is highly improbable.
A more probable scenario is that bacteria of the microbiota or the E. coli inside our body exchange DNA with E.D. Frosti. To avoid this, we added a biobrick with DNase activity in our project. We chose the DNase activity as it degrades the DNA without cell lysis and also shuts down its own induction. The degraded DNA stays in the cell, which makes the exchange of DNA with other organisms less probable to occur. The bacteria will be released into the environment once the proteins have been formed and cell death has occurred. Cell debris are found all over the world and we believe that the cell debris won’t cause any problems, because the entire DNA is already degraded and no toxic compounds are found inside the cell.
6. Conclusion“How safe is safe enough?” We will never know the answer to this question. The only thing we can do is to try to protect ourselves and the environment as much as possible and in case something goes wrong, try to minimize the impact of the outcome. Until now reports of major health problems with genetically manipulated microorganisms are scarce, if not absent. We believe that this is the result of the extensive risk analysis that are made beforehand and the necessary precautions that are taken while experimenting. We intend to make an organism that freezes and defreezes water. It is important that this organism doesn’t overgrow the environment or have toxic properties. We have taken enough safety precautions in our project to minimize the potential negative effects.
The discussed safety problems can be used by people who are opposed to genetically manipulated organisms to scare lay people (people who don’t know anything about this subject). It is important to explain them that the probability of these scenarios happening is very low. How to do this is further discussed in bio-ethics and education section.
7. Additional safety remarks due to comments given on the regional Jamboree: prevention of horizontal gene transferIf our suicide mechanism would fail e.g. because of a mutation, than the probability of the plasmids being transferred into other organisms in the neighborhood would increase. This process is called horizontal gene transfer. The future plans of using E.D. Frosti commercially for the induction of ice crystallization in ice rinks and prevention of ice formation on the roads requires the release of a huge amount of cells into the environment. This brings the E.D. Frosti in contact with other bacteria and the survival of E.D. Frosti could lead to horizontal gene transfer. If we take into account that these bacteria are well adapted to the natural habitat, we can assume that the probability of the apocalyptic scenarios occurring would increase enormously after horizontal gene transfer. To minimize the chances of these scenarios, we searched for new safety mechanisms to prevent horizontal gene transfer and additional safety mechanisms to ensure the cell dies when released into the environment.
Horizontal gene transfer
While brainstorming for a mechanism to prevent horizontal gene transfer, we concluded that the best solution would be to avoid the use of plasmids. Horizontal transfer occurs when a plasmid is transported from one bacterial cell into another cell. If all biobricks could be incorporated into the bacterial genome, the probability of horizontal gene transfer would drastically decrease.
The mini Tn7 biobrick of the Sevilla iGEM team 2011 is an excellent example of this mechanism. We could insert the AFP- and INP-genes in a mini Tn7 plasmid. The mini Tn7 would transfer these genes into a specific site of the E.D. Frosti genome. The remaining residues of the plasmid would be broken down and horizontal gene transfer would be disabled. Besides preventing horizontal gene transfer, this mechanism gives us also the possibility to remove the genes that are coded for antibacterial resistance.
Additional safety mechanisms
Biological containment is the most essential safety measure. Scientist can do this physically by keeping the bacteria in the lab or factory (industrial productions). But we plan to release E.D. Frosti into the environment, so as solution to biological containment we have inserted a cell death mechanism. We cannot control mutations in the genes coded for cell death, but we can ensure that E.D. Frosti dies by taking additional safety measures. One of the measures can be deleting genes that code for the production of several vitamins. The bacteria would only be able to survive when we add those vitamins to the medium. This way, even if the ceaB mechanism fails, the probability of E.D. Frosti surviving in the environment is heavily reduced.
Adding more safety mechanisms equals reducing the probability of survival of E.D. Frosti in nature. Reducing the probability to zero is impossible. So the question remains: how safe is ‘safe enough’?
- Safety and laboratory practice link, 05/07/2011
- Schmidt M., Kelle A., Ganguli-Mitra A and Vriend H., (2009) Synthetic biology - 2009 The technoscience and its societal consequences, 81-97 (chapter 6)
- Dr. coli, link, 06/07/2011
- Bacilla filla, link, 06/07/2011
- Hickman Jr. C., Roberts L., Keen S., Larson A., I'Anson H., Eisenhour D., (2008) Integrated Principles of Zoology, McGraw Hill Higher Education 14th edition
- Biotechnology Program under the Toxic Substances Control Act (TSCA), (1997) Escherichia coli K-12 derivatives final risk assessment, 12/07/2011
- University of Wisconsin – Madison, E. coli genome project, link
- Hoffman F. P. and Schrag P. D., (2002) The snowball Earth hypothesis: testing the limits of global change, Look for another paper about global warming, Blackwell Science 14, 129-155
- National Academy of Sciences, National Academy of Engineering, Institute of Medicine, National Research Council, (2008) Understanding and responding to climate change - Hightlights of National Academi es Reports
- Windows to the universe, link, 11/07/2011
- Wu Z., Qin L. and Walker V. K., (2009) Characterization and recombinant expression of a divergent ice nucleation protein from ‘Pseudomonas borealis’, Society for general microbiology 155, 1164-1169
- Goodnow R.A., Katz G., Haines D.C. and Terill J.B., (1990) Subacute inhalation toxicity study of an ice-nucleation-active Pseudomonas syringae administered as a respirable aerosol to rats, Toxicology Letters 54, 157-167
- Wrzaczek M. , Brosché M., Salojärvi J., Kangasjärvi S., Idänheimo N., Mersmann S., Robatzek S., Karpiński S., Karpińska B. and Kangasjärvi J. (2010) Transcriptional regulation of the CRK/DUF26 group of Receptor-like protein kinases by ozone and plant hormones in Arabidopsis , Biomedcentral Plant Biology 10
- Mini Tn7 Biobrick (iGEM team Sevilla 2011),link, 13/10/2011