As a new team to iGEM, we were concerned about safety and potential dangers that synthetic biology, and particularly our project, could create and how they are prevented in the competition. That's why we interrogated ourselves not only on risks on a strictly biological point of view, but also in a more ethical and dual-use oriented way of interpreting what is produced, its misuses and how to prevent them.
This page is a sum-up of all the discussions we had about safety,
Biological Risks
Used Strains
Being aware of the safety level of the bacterium you work with is one of the most important preventive measures for a biotechnologist. Indeed, we discussed safety risks to reduce them as much as possible, both for our research and in potential industrial applications of our solution.
Developing our project we used two strains: Rhodococcus opacus PD630 and Escherichia coli DH5α.
We found documentation certifying the biosafety level needed to handle Rhodococcus opacus PD630 and confirmed that BSL1 facilities were sufficient (https://bacdive.dsmz.de/strain/10996#ref415). We also contacted the iGEM safety team to clear all doubts about its applicability in an iGEM project, and we were approved without the need of a check-in form for it. The other strain used in our experiments (Escherichia coli DH5α) is widely used in labs and in various iGEM projects. It has a BSL1 level and therefore didn't require any additional approval procedures.
Furthermore, as explained more in depth in the human practices section, we discussed with the infectious disease specialist Dr. Giovanni Dolci how Rhodococcus strains are treated in case of infections, and their pathogeny. We learned that Rhodococcus strains aren't prone to become human or mammal pathogens, with the sole exception of Rhodococcus equii, which is mainly responsible for infections in horses and other domestic mammals. Cases of spillover to humans are sporadic and mainly recorded on immunodeficient patients. In any case, Rhodococcus strains proved to be susceptible to a huge variety of the most common antibiotics used for therapies. This is why we consider Rhodococcus opacus PD630 to be safe for application, both in lab and in a potential industrial application.
Good Practices for Biosafety
Safety precautions in the lab act towards two main goals: protecting the operator, and protecting the environment to avoid releases.
To protect the operator, Personal Protective Items (PPIs) like lab coat, nitrile gloves and safety goggles were provided. Also, long trousers and closed footwear were always worn, and it was forbidden to work alone in the lab.
To avoid releases, biological materials containing live organisms were always treated in a sterile environment, both a BSL 1 safety hood or a sterilized workstation with Bunsen burners. Furthermore, all biological liquid waste was treated with sodium hypochlorite before disposal, and solid waste was treated accordingly to Italian regulations, and disposed in double sealed containers.
Dual Use of Our Chassis in an Open-Source Environment
One of the main reasons of concern in our project was given by its chassis nature: our goal is to facilitate gene integration. This raises the problem of what genes will be inserted in the chassis, since there is virtually no control on what may be inserted.
We interrogated ourselves on the responsibilities of scientists that may use this idea and replicate it. We concluded that open-source synthetic biology gives less control on what other labs will perform, heavily relying on the ethics of scientists who will use the technology. At the same time, open-source technology has the immense advantage of allowing a huge number of labs to work on managing real problems, creating new solutions based on ones that already proved to work. This is why we think that if the application follows synthetic biology rules, it will result in no substantial harm.
Anyways, the fact that we feared the most is that potential dual use of the technology is not public and can be performed in secret, also by using the technologies developed here. We have virtually no control on this phenomenon, but we considered that the challenges faced during this project are a bottleneck that makes the development of new chassis with cre/lox systems hard and time consuming. This allows the chassis developer to have control on who will receive it, since it is way easier to share the strain rather than creating it by scratch. Limiting the availability of the chassis to only trusted labs is a way to control that the technology will be applicated safely, preventing dual-use.
Propagation Containment of a Potential Application
During our human practices meetings, we had the opportunity to meet Davide Beltrame, research and development team member at the Acque del Chiampo S.p.A. wastewater treatment plant in Arzignano, Vicenza, Italy. Among the possibilities discussed there, better described in the human practices section, we hypothesized what should change to implement our bacteria in standard plants.
We identified two main critical points. First, the addition of a new member to the already established community may lead to changes in the normal degradation process. This problem can be solved with preliminary studies. Secondly, some smaller wastewater treatment plants may need to implement an additional step of bacteria elimination to not release the genetically modified organism in the environment.
Antibiotic Resistance in Application
During our conversation with Dr. Giovanni Dolci, we also discussed how antibiotic resistance conveyed by our bacteria may affect the environment and potential treatments in case of infection.
The peculiarity of our chassis is that it will integrate all genes and new parts added in its own genome. This implies that antibiotic pressure is not needed to keep functionality for a possible application, being necessary only for selection of the chassis of our interest. This means that every correctly modified bacterium will have successfully integrated 2 antibiotic resistance genes: one for thiostrepton, and another for kanamycin. These two genes will not be in plasmids, highly decreasing the chance of horizontal gene transfer to other bacteria. Anyways, in case of infection caused by our Rhodococcus opacus PD630 chassis (which, as stated in the first section of this page, has never been registered as pathogenic) would still be susceptible to most of the antibiotics currently in use
Informatic Risks
Design of Harmful Constructs
The same features that make our tool useful for bioremediation applications (prediction of optimal pathways, codon optimization for our Rhodococcus opacus PD630 chassis, downloadable assembly-ready sequences) could, in theory, be used to design pathogenic or toxic genes. This raises the possibility of misuse of our software to produce toxic proteins/intermediates, virulence factors, or antibiotic resistance determinants.
Additionally, there is also the concern that the software could be useful to circumvent natural genetic barriers. Normally, differences in codon usage or inefficient expression reduce the impact of foreign DNA in microbial systems. By optimizing for efficient expression, our tool lowers these limitations, which, in the wrong context, could make harmful constructs more viable.
Given the high flexibility of our solution, we cannot in principle ensure that construction of harmful pathways is avoided. However, an important consideration to be made is that our pathways prediction algorithm is designed exclusively for reactions leading to pollutant degradation. A user aiming to design synthetic pathways would only be able to rely on publicly available resources such as KEGG, which are already accessible online independently of our project.
Still, we acknowledge the possible risks explicitly and document the intended, beneficial applications of the tool very clearly, appealing to the common sense of users. Although not implemented yet, another potential safety control could be the automatized screening of genetic constructs against databases of toxins, virulence factors, and clinically relevant resistance genes. Such a feature could alert users when a designed construct contains sequences with known harmful potential, or even block the download of such designs.
Adaptation to Unintended Hosts
Although our system is specifically tailored to Rhodococcus opacus PD630, its scalability implies that malicious actors could modify the code for use with other bacteria, including those that are not considered safe.
The open-source nature of iGEM projects, while increasing accessibility on one hand and being a strength for collaboration, can also be considered a weakness if technologies are taken out of context and used without appropriate safety measures. However, we have taken steps to reduce the practical risk of such misuse: our published software focuses on parameters and part libraries optimized for the Rhodococcus chassis (e.g., codon tables, regulatory element choices and assembly conventions), and we do not provide curated part collections for other, higher-risk organisms.
Chemical Risks
Mutagen Chemicals
The only mutagen chemical we implemented during the project is ethidium bromide. We are aware that this reagent may be more dangerous than other DNA staining molecules, but since concentrations used during our experiments are always lower than the LD50 in rats, we decided to use it anyways, at the condition of treating it with maximum care, as if it were toxic.
Flammable Chemicals
The main flammable chemical we used is ethanol, which we implemented for sterilization and cleaning of surfaces. The main source of risk was when Bunsen burners for sterile environment were active. In these situations, ethanol containers were kept at least 2m away from the flame, and equipment was sterilized with ethanol prior to flame ignition or by exposing them to the flame if possible.
Chemicals Not Listed in the Safety Form
Since HELMET project was still trying to find new ways to solubilise lipase, some new solvents were implemented after the deadline for safety form submission. Here is a list of those which were listed as dangerous:
- Isopropyl alcohol, CAS 67-63-0, flammable, mildly toxic
- Triton X-100, CAS 9002-93-1, harmful if swallowed, toxic to acquatic life
- Dimethyl sulfoxide, CAS 67-68-5, irritant
- Hydrochloric acid, CAS 7647-01-0, corrosive
- Tris(hydroxymethyl)aminomethane, CAS 77-86-1, irritant
- Para-Nitrophenyl palmitate (4-NPP) CAS 1492-30-4, irritant, toxic if ingested, flammable, harmful if inhaled
- Disodium hydrogen phosphate (dihydrate) CAS 10028-24-7, irritant
- Sodium dihydrogen phosphate CAS 7558-80-7, irritant
We prevented any harmful contact wearing gloves, lab coat, long trousers, closed footwear and safety goggles at any time in the lab.