Team:KULeuven/Ethics/Robotics/Robotics

Biological robotics
In more than one way, one easily recognizes some of the preliminary features of nanorobotics in Dr. Coli. Nanorobots are described as devices that can manipulate objects at the nanometer scale. This is the same order of magnitude as the size of atoms and molecules. Indeed, Dr. Coli is a kind of robot, that can deliver drugs where and when they are needed inside the human body. However, we would exaggerate saying it is a nanodevice. The typical dimensions of an E. coli cell are micrometers (10^-6) rather than nanometers (10^-9). This does, however, not stop us from treating Dr. Coli in terms of robotics.

Three laws of robotics
In his short story “Runaround”, published in 1942, the Russian writer (and professor in biochemistry!) Isaac Asimov introduced his “Three laws of Robotics”. They are: "A robot may not injure a human being or, through inaction, allow a human being to come to harm." "A robot must obey orders given to it by human beings, except where such orders would conflict with the First Law." "A robot must protect its own existence as long as such protection does not conflict with the First or Second Law." We will think about Dr. Coli here, using the “Three laws of Robotics” as a framework for further reflection.

In “Do No Harm To Humans: Real-life Robots Obey Asimovs Laws”, the writers point to an interesting dilemma in working with robots. On the one side, it is easy to make robots do all kinds of stuff, thereby obeying the laws of robotics, if they are far away enough from mankind. On the other side, as long as robots work slow enough, they say, the robots cannot violate the laws. "The trade-off between safety and performance is the name of the game in physical human-machine interactions. (6)"

This understanding makes our reflection on Dr. Coli safety issues not easier. On the contrary, it makes it even more difficult. Dr. Coli does work the closest to man one can imagine: inside his body. Even more, it is practically impossible, and indeed even heavily unwanted, that the speed or efficiency of Dr. Coli is lowered.

This second understanding imposes a lot of responsibility to the creators of Dr. Coli. We will have to go through the three laws of robotics and discuss in detail how we can make sure that Dr. Coli does not violate a single one of them.

First Law
"A robot may not injure a human being or, through inaction, allow a human being to come to harm." Asimov’s ideas about the three laws of robotics date from the first half of the twentieth century. In these times, no one was actually concerned about environmental health, global warming and fossil fuels. It is therefore no surprise that the first law of robotics needs critical review. We now are concerned about environmental health, global warming and fossil fuels (or the lack thereof). In this view, we can, and should in fact, expand the first law of robotics. Not only is a robot—or in our analysis, Dr. Coli—prohibited to allow a human being to come to harm, it should also be prohibited to allow any other part of nature to get harmed. Under any part of nature, we mean the environment in its broadest sense. It is indeed extremely important not to create such things, that can interfere with nature and, by doing so, harm it.

This remark is a topic of general concern and is one of the issues that has been discussed maybe more than all other synthetic biology topics. It is in this context that a lot of discussion resembles the issues discussed during the Asilomar conference of 1975 on genetic engineering. During that conference, organized by Paul Berg, the world’s leading scientists heavily discussed the safety of recombinant technologies. That science was brand-new in those years and raised a lot of questions about whether it was safe to alter DNA.

Much of the concerns resulted in regulations and agreements that are very common today (for instance biological and physical containment) or are either still actual for synthetic biology. Therefore, we will focus on Dr. Coli and reformulate the question concerning Dr. Coli and the First Law of robotics as following: "How can we make sure that Dr. Coli does not harm any part of the enivronment?" We notice that we explicitly formulate the broad meaning of the word environment. This environment consists of three parts: the patient, other human beings and the non-human nature.

Patient
Taking the built-in safety mechanisms of Dr. Coli (see project description page) into account, the only way for him to bring harm to the patient is when he is given to a patient that is ill, but has not the illness the specific Dr. Coli is specialized in. Another possible danger is that Dr. Coli is given to or taken in by a non-infected patient. That way, Dr. Coli’s memory is in the 0-state when initially given to the patient and will not be activated, as there is no input signal. This is indeed a dangerous situation, for this situation will lead to uncontrolled cell growth that can only be stopped by using the right antibiotic treatment. This is to be avoided, as antibiotics are now, more than before, to be used with care to avoid the development of multiresistant strains.

We can stress the importance of medical doctors here. Dr. Coli is indeed no different than other drugs in that it is dangerous to give the medicine to a healthy person or a person with a different problem. However, it is a very easy and irresponsible attitude to completely pass on the safety issues about Dr. Coli to the medical staff. We cannot hold them fully responsible for possible problems with Dr. Coli.

Other human beings
Dr. Coli is a bacterium that resides in the colon. Keeping in mind that the average household has not the same kind of safety regulation as the average biotechnological lab, we can be sure that Dr. Coli will be transmitted to people coming near the patient. This should, however, not pose a problem. In the case the patient is ill, Dr. Coli’s memory will be in a switched on state once it leaves the patient. This means that as long as no input signal is present, Dr. Coli will count to the moment he will kill himself.

In the case the patient is not ill at all, or has a different illness, Dr. Coli can in fact spread in human beings, other than the patient.

Non-human nature
Also here, we can divide the risks into two classes. On the one hand, Dr. Coli is dangerous for the non-human nature after he has passed the intestinal system of the patient. On the other hand, Dr. Coli is dangerous for the non-human nature prior to administration to the patient. One can for instance spoil some Dr. Coli solution or a leak in growth reactors in the pharmacological factory can originate. However, such accidents can be monitored appropriately and heavy consequences can be prohibited by means of antiseptic agents.

Second Law
"A robot must obey orders given to it by human beings, except where such orders would conflict with the First Law." We can think about this law the following way. The orders given to Dr. Coli is to heal the patient when and where necessary. Furthermore, Dr. Coli is instructed to die after a period of absence of illness signal. This is encoded in its “software”, its genes. So the only way for Dr. Coli not to obey his orders is to mutate. There is no way of completely preventing this. Regular sequencings to check whether the sequence is still consistent and changing to a fresh culture derived from an early seed culture should become part of every production plan for Dr. Coli and become a standardized procedure. An other thing that can happen to Dr. Coli is the loss of the plasmid containing the working mechanism of Dr. Coli. This can, however, be more or less overcome by integrating the plasmid in the chromosome.

Third Law
"A robot must protect its own existence as long as such protection does not conflict with the First or Second Law." This means that, based on easy logic, we can alter the Third Law to the following statement: "If the existence of a robot conflicts with the First or the Second Law, it may not defend its own existence." A conflict with the first law can, for instance, occur when Dr. Coli escapes the biological and physical containment imposed on it. We will provide a more elaborate workout for the issues with Dr. Coli ending up in the environment. We quote the Summary Statement of the Asilomar Conference on Recombinant DNA Molecules: "Containment of potentially biohazardous agents can be achieved in several ways. The most significant [...] is the use of biological barriers. These barriers are of two types: 1) fastidious bacterial hosts unable to survive in natural environments, and 2) nontransmissible and equally fastidious vectors [...] able to grow only in specified hosts. (7)"

One of the best possibilities to put up a biological barrier, is replacing an essential gene by the Dr. Coli construct. A conflict with the second law is less feasible, as Dr. Coli does not have a choice but to follow his genes. Mutating or dying are the only things Dr. Coli can do to escape the directions given to him by us. As we said before, screening methods can monitor any abnormal behavior and easy and quick measures can be taken.

Dr. Coli safety alterations
It has not escaped our notice that if we ever want to implement Dr. Coli, more safety alterations have to be added. First of all is E. coli certainly not the best bacterium available to apply the system in human beings.

One of the best bacteria to carry out the function of Dr. Coli, would be Lactobacillus. Lactobacilli are gut-friendly gram-positive bacteria, widely used in probiotics. As Lactobacilli are already present in the gut, they are completely harmless. An important feature of Lactobacillus is the fact that it’s an immobile bacterium. So once arrived in the gut, it stays there. The risk of spreading and contamination of the environment is thus minimized.

To reduce the risk that activated Dr. Coli can reside somewhere we don’t want it to be, we can place a second AND-gate at the beginning of our system. This double input can only activate the system if both signals are available. The signal of sickness is one input, the other input can be a substance we add to the diet of our patient. We have to make sure that the second input is very specific and is something normal people don’t take in.

Another point we have to consider, is the fact that the second input has to be an essential substance for the cells. So if there’s only one input available, Dr. Coli dies. Up to now, Dr. Coli’s active elements are gathered on plasmids. This is, however, far from the best method for making cells for administration in the gastro-intestinal system, where they can make a therapeutical agent. An improved version of Dr. Coli should for instance contain the active elements in the chromosome.

Researchers at the VIB (Flanders Institute for Biotechnology) already demonstrated a mechanism of this kind. They engineered a Lactobacillus strain to produce IL-10 in the treatment of Crohn’s disease. To be allowed to do studies in clinical phase with testing on humans (they already arrived at that point), they needed a perfectly safe mechanism that allows the Lactobacillus to live in the patient, but inhibits growth outside the body. The mechanism used by the VIB research group made use of a Lactobacillus strain in which the thymidylate synthase gene (thyA) has been replaced with a gene producing a pharmacological agent (in their case, IL-10). This means that the bacterium will die in response to thymine and thymidine starvation (Ahmad, S. L, Kirk, S. H. and Eisenstark, A. (1998) Thymine metabolism and thymineless death in prokaryotes and eukaryotes. Annu. Rev. Microbiol. 52,591-625.).

Finally we want to emphasize that our Dr. Coli is a proof of principle and that the concept is far from ready to be implemented in human beings.