---

Our problematics

---

A specificity of our project is that we did not start its construction from a single problematic but we actually started from a multiplicity of problems. For examples:

• How to produce aid bandings with antibiotics?
• How to produce paper where we can print electronic circuit boards?
• How to create new fluorescent fabrics?
• How to create packaging exhaling the smell of the food inside?
• How to manufacture feminine protection preventing toxic shock syndrom?

and so on

This forced us not to think about a specific real world problem but to tackle a technical lock common to all these project realizations: functionalizing cellulose.

Each questions could have been addressed using a dedicated physico-chemical or even synthetic biology solutions. However, we considered this an awkward way to solve such a generic problem and we decided to search for a generic and more elegant solution. This is how Cerberus has been brought to life: instead of one construction for each question, we deeply brainstormed about creating a solution for all the questions.

In the specifications, the most important features we sought for were the genericity and versatility. We found rapidly what molecules should be grafted to cellulose, but they were of a very different nature to be associated through a common property: graphene, antibiotics, carbon nano-tubes, fluorescent proteins or chemicals, magnetic beads, etc.

Emergence of the Cerberus platform

---

It soon appeared that there were three main properties important for us: (i) to bind cellulose, (ii) to bind organic compounds, (iii) to bind inorganic molecules. This is how we designed a three headed protein where each head answers to one of these specific questions (and how we choose Cerberus as a name for it!).

The cellulose head was the easier to design. Cellulose Binding Domains have already been successfully used by fellow iGEMers (for example, Imperial 2014 or INSA Lyon 2016). This required a little sequence optimization but it was easy.

The next head was the streptavidin domain. Here we had to choose between the classic tetrameric streptavidin and a monomeric streptavidine unsuccessfully used by iGEM UGent Belgium 2016. We tried both. We also added a TEV protease site to add the possibility to release the molecule associated to streptavidine, even if we did not try it in the course of the iGEM competition.

The last head was the trickier. To bind non organic molecule to a protein, we had to search in state-of-the-art synthetic biology concepts, at the interface with chemistry. We ended up with choosing unnatural amino acid incorporation in the Cerberus Platform and the use of click chemistry to create covalent bound between inorganic or semi-inorganic compounds and our protein.