After assembling our team, we spent the first two months in brainstorming sessions, researching potential topics for this year’s competition. A team member presented us the field of bionic prosthesis, and we were all shocked by the number of people who were suffering from an amputation. His first idea was to use the capacity of bacteria to drive an electrical current to amplify the signal between the patient’s nerves and electrical sensors linked to a prosthesis. This way, redirection of nerves wouldn’t be necessary and the patient would be able to accomplish more natural actions.

When we started to look more in details the field of prosthesis and implant, we realized that along with device loosening or malfunctions and foreign-material reactions, infection remains the most serious problems encountered with surgical implants. We quickly learned that biofilm formation is common to all types of implanted foreign-body infections. Indeed, the high susceptibility of implanted devices to infection is due to a locally acquired host defense deficiency. Thus, this persistence at a specific site is mainly because of the rapid formation of a biofilm, which is resistant to host defense and antimicrobial agents as a result of reduced access and diffusion characteristics within it. We then decided to integrate this aspect inside our project: while favoring the growth of the nerve and the conduction of a signal through a bacterial interface, we could also diminish the risk of infection. At first, we wanted to find a system to directly kill the S. aureus, but after talking with Dr. Jean-Marc Ghigo (Genetics of Biofilm Unit, Institut Pasteur), we realized that the biofilm configuration would give us some hard time and require extra steps to manipulate. We decided to shift our goal, and rather than killing S. aureus, we would limit the virulence of the bacteria and restrict its ability to form a biofilm. This could be achieved by subverting the quorum sensing of the pathogenic bacteria and blocking the signal. In this way, S. aureus could be handled by the host’s immune system and by the patient’s doctor with a normal dosage of antibiotics.

In order to learn more about the significance of these issues in the medical field, we interviewed many professionals working either directly with patients, industries working on high tech metallic or ceramic implants, and amputees through the contact of ADEPA « Association for the Defense and Study of Amputated People ». Through all this work, we were invited for many tours, first at the European Hospital George-Pompidou where we had the chance to observe Dr. Benjamin Bouyer M.D., a lumbar rachis surgeon, during surgery to see the procedures put in place to diminish the risk of infection during the integration of an implant inside the body.

The first thing we needed to do was to narrow down the pathogenic bacteria we would work on. After researching in the literature and talking with experts in the fields of prosthetics, we found out that Staphylococcus aureus and Staphylococcus epidermidis are the two most frequently found bacterial infectious agents on implants. We decided to use the biobrick designed by the iGEM team LMU_Munich 2014, who worked on a genetic circuit enabling Bacillus subtilis to actively detect Staphylococcus aureus.

Our bacterial chassis of choice, which would inhibit the quorum sensing of Staphylococcus aureus, needed to be a bacterium naturally occurring inside the human body. Indeed, we didn’t want to introduce a new strain inside the human body. The chassis also needed to have a good secretion system, as well as a good tolerance for the immune system. We narrowed down our choices between Bacillus subtilis, which the iGEM LMU_Munich 2014 worked on, and Escherichia coli, naturally living inside the gastrointestinal tract of humans with mostly no problem except for some strains. Because we wanted to improve their construction and because of convenience inside our lab, since we already had competent E. coli cells, we decided to use this strain. We optimized our sequences for an E. coli expression, and adapted them to its secretion system by adding export signals. This way, our construction could be broadened and used in a different type of bacterium. To replicate our plasmids, we used Escherichia coli DH5 alpha and to secrete our proteins, Escherichia coli BL21De3 pLys S.

When conceptualizing our proof of concept device, we decided to design a microfluidic chip to simulate the actions that would occur inside the patient’s body. The chip would be capable of measuring the neuronal signal as well as the conductivity of the biofilm, letting us know whether our system would work correctly inside a prosthetic. After talking with many professionals such as Dr. Heng Lu (ESPCI) and Dr Ayako Yamada (ENS, Ecole Normale Supérieure, Paris), we first designed a PDMS microfluidic chip with a vitreous carbon electrode to measure the signals. After talking with Dr. Catherine Villard (Institut Curie, Paris) and Dr. Frederic Khanoufi (University Paris Diderot—ITODYS), an electrochemist, we realized this type of electrode was highly sensitive, perhaps too sensitive for our type of device, and not adaptable to the size of our microfluidic chip. Guided by their advice, we switched to gold electrodes and started the fabrication process of the different type of chips at the Pierre Gilles de Gennes Institute. After attending the iCOE 2018, we learned about PEDOT (poly(3,4-ethylene dioxythiophene) polystyrene sulfonate ), and its conductive properties. As we also wanted to confine our bacteria so they would not harm our neuronal cells, we partnered with Sterlitech, and tested nanoporous polycarbonate membranes coated in gold as well as nanoporous alumina oxide membrane coated in PEDOT: PSS, PEDOT: CL and PEDOT: TS. This way, we could still measure the neuronal signal and the conductivity of the biofilm while protecting the cells from getting eaten by the bacteria.

When we started to think about the scientific aspect of our project, we also started to design and think about how our biofilm would integrate into a physical medical device. First, we wanted to design a full prosthesis. We realized that it wasn’t the core of the problem. Indeed, the technology missing in this field was the actual interface between the prosthesis and the osseointegrated steel/titanium/ceramic stem inside the human body, limiting the field of bionic prosthesis. We decided to focus on this interface and started to talk with Dr. Benjamin Bouyer, a Lumbar rachis surgeon, on how we could deposit the biofilm on or in the osseointegrated stem. We realized that the stem would be in direct contact with the surgeon meaning that if we coated the whole structure with our biofilm, it would probably be stripped off, and get deposited on the gloves and contaminate other areas. We also spoke with one member of the board of directors of ADEPA, “Association de Défense et d’Etude des Personnes Amputés », which translates to « Association for the Defense and Study of Amputated Persons” who is himself an amputee. He gave us great advice on the designing phase of this interface and raised the issue of the socket causing excessive sudation and discomfort for the patient. After meeting with experts from I-CERAM (Ceramic medical devices company, Limoges, France) and the CERAH (Center for Studies and Research on the Equipment for the Handicapped), we learned more and more about the different materials used in the making of prostheses. At this point, we realized that doing a full stainless-steel interface would stimulate the growth of bacteria in a biofilm structure. Therefore, we decided to switch and do the part in direct contact with the patient in ceramic, knowing that we still have the same problem that I-CERAM and the CERAH were facing currently. Indeed, steel doesn’t last as much as ceramic, which could be a problem when creating an interface composed of both materials. We knew then that the patient would probably need corrective surgeries to fix his osseointegrated stem. Having considered these parameters, we then decided to start modeling our prototype by integrating those different aspects as much as possible. Because we wanted it to be cost-effective and injection-moldable, we built our current prototype in ABS, a thermoplastic polymer that could be 3D-printed. We ordered the electronic parts composed of a charger, battery, and amplifier, and assembled it into its current state as our POC.

While we were doing adjustments to our design in the lab, our team of jurists also researched how our device could be integrated into society given the current political and economic landscape in France. The use of GMOs, for the environment, food industry or medicine, is highly regulated in France and Europe. It is limited to design animal models of diseases and to produce large quantities of molecules for the pharmaceutical industry. The use of genetically modified bacteria inside the human body is not the subject of specific laws, which made our research on the subject difficult. We focused our research on medical device’s regulation in order to see if our project could be marketed. At first, it appeared almost naturally that our project was a medical device, but as it has to be understood as a legal term, it responds to a specific definition. Working on medical devices’ regulations helped us to better define our project, classify it and analyze our project in the scope of European regulations. We also tried to cover some legal questions related to our project and tried to find answers and other legal system such as the American one.

All these feedback from scientific experts, association members representing patients, and physicians/surgeons, allowed us to develop our project NeuronArch to what it is today. Without them, our project would not have been the same, and we really want to thank them for that.