Cellulose: a material with endless possibilities

With approximatively fifty billion tonnes produced yearly, cellulose is one of the most abundant polymers on earth¹. Cellulose is a polysaccharide made of D-glucose units linked together in a linear chain by β-(1-4) bonds. It is mainly produced by plants, but also by bacteria such as Gluconacetobacter hansenii. Nowadays, cellulose and its derivatives (nitrocellulose, cellulose acetate, methyl cellulose, carboxymethyl cellulose, etc.) are among the first materials that have been exploited in the industry (wood, cotton, textiles, paper, electronics and biomedical devices²), and they now represent significant economic issues.

However, today, we mostly use cellulose that is extracted from plants and their wastes. But because it branches with other compounds such as hemicellulose and lignin, it must go through harsh chemical treatments to finally be considered “pure”³. These purification processes cost a lot in terms of energy, money and time. One solution to this problem could lie in the use of bacterial cellulose. Some acetic acid bacteria naturally produce cellulose as biofilm to protect themselves. This cellulose is purer than the one produced by plants; it shows also interesting properties compared to the other one (unique nanostructure, high water holding capacity, high degree of polymerisation, high mechanical strength and high crystallinity³) which proves its potential as a promising material and alternative to plant cellulose for the future. Sadly, due to its high cost and low-yield production, bacterial cellulose is not yet widely spread in industries.

By functionalising cellulose, we aim to demonstrate how this widely used material can be more than a simple packaging or textile fibres. By combining it to Cerberus, cellulose can become a source of information and bring new opportunities for industrial applications.

Genesis of our project, Cerberus

When brainstorming about the project, we had these different ideas and they all revolved around functionalising several materials. We then realised it could be of great use to have one material that we could functionalise at our discretion. That is how we decided to work on cellulose, especially bacterial cellulose, in hope of developing its use by adding a whole range of exciting possibilities. From conductive paper to anti-infection tissues, applications domains are aplenty (medicine, textile, stationery, etc). The major bottleneck is the complexity to conjugate bioactive molecules to cellulose.

To circumvent this limitation and enable a wide range of chemicals to be fixated to cellulose, we designed a three headed linker protein named Cerberus (in reference to the mythological dog). Cerberus is based on the fusion of three fixating protein structures representing the three heads of the system. The first head is a protein domain of the type 3 Carbohydrate Binding Module (CBM3) family to bind cellulose. The second is one of the strongest linkage systems of the living realm, streptavidin, with its high affinity for biotinylated compounds. The last of the heads features a non-natural amino acid, 4-L-azidophenylalanine, catalysing covalent bonds by click chemistry. Therefore, the versatility of our thought-out linker will allow a large variety of organic and inorganic molecules to conjugate with cellulose.

One head to bind to cellulose, the CBM3a

The first head is the core of our linker, Cerberus. We chose the Carbohydrate Binding Module type 3a (CBM3a) to bind to cellulose. This domain is found in the cellulosome of Clostridium thermocellum⁴ and its four aromatic amino acids make it the perfect candidate to bind to cellulose through a dipole-dipole interaction between the CBM3a’s aromatic amino acids and the cellulose’s glucose rings.

Our CBM3a features a long endogenous linker on each end, to which we can attach other structures to functionalise cellulose.

One head to bind biological macromolecules, the streptavidin

The second head of Cerberus is streptavidin, which have been added to cellulose N-Terminus. Streptavidin is a 60 kDa homotetrameric protein, isolated from Streptomyces avidinii, and it is well known for its interaction with biotin and biotinylated proteins. With a dissociation constant of 10-13 M, the interaction between streptavidin and biotin is the strongest non-covalent one known in Nature⁵.

The sequence of streptavidin wild type is used in fusion with the CBM3a. This streptavidin forms a complex with three other streptavidin, hence the unmatched affinity for biotin since it can bind four molecules of biotin per tetramer. We decided to use two versions of streptavidin: the native form which is tetrameric because we have data on its interaction with biotin, and a monomeric version called mSA2⁶ because we feared that the tetrameric streptavidin would cause agglutination of our protein Cerberus. In case we needed to release the streptavidin linked to our function into the surrounding medium, we also included a TEV-protease site to cut off this section.

By adding a 15-residue AviTag, we managed to biotinylate our proteins with the help of BirA⁷. For this in-vivo biotinylation, we have transformed our chassis with a plasmid containing the protein BirA.

One head to bind inorganic molecules, the 4-L-azidophenylalanine

To expand the versatility of our system, on the C-Terminus of the CBM3a, an unnatural amino acid called 4-L-azidophenylalanine was added. Its structure is like that of a phenylalanine, except that it presents an azide, or N-three, function on its tip, as the name would suggest. When presented with an alkyne function, and in particular dibenzocyclooctyne (DBCO), a reaction called click chemistry allows us to covalently bond these functions to our Cerberus protein in a spontaneous and irreversible way.

The strategy used to incorporate this unnatural amino acid into our protein is to bring an orthogonal pair of tRNA and aminoacyl-tRNA synthetase with our plasmid⁸. The tRNA will recognize the amber codon TAG and this codon will not stop the translation as it normally does. Instead, it will allow the incorporation of our unnatural amino acid, the 4-L-azidophenylalanine.

There are two types of click chemistry that can be done on 4-L-azidophenylalanine to functionalise our Cerberus. The first one is the copper-catalyzed azide-alkyne cycloaddition reaction (CuAAC). It takes place as its name suggests between the azide group of the unnatural amino acid and an alkyne group; the reaction is catalysed by Cu2+ ions. The main problem of this method is the cytotoxicity of copper on cells limits its use in assays that take place in the cellular environment. The second type of click is the strain-promoted azide-alkyne cycloaddition reaction (SPAAC) which proceeds without the need of a catalyst⁹.

For our experimental plan, we decided to use both types of click chemistry, but we managed to only use the SPAAC during our time in the lab.

Our system

For further details about our system, please see our Design page.