Demonstrate

---

Introduction

---

Our entrepreneurial study allowed us to evaluate the diverse possibilities that our Cerberus protein can offer in terms of functionalities for bacterial cellulose. It also permitted to feature the versatility of our protein thanks to its three link sites: a binding molecular platform through the capability of CBM3a to interact with cellulose, and two functionalizable linkers thanks to (monomeric) streptavidin and the unnatural amino acid (UnAA) AzF moieties. This section is dedicated to display that, in fact, we reached our goals of fixation of molecules to bacterial cellulose. We have divided the full procedure in three steps.

Validation of the Three Binding Heads

---

We designed three proteins, each allowing to validate one binding head. We designed first Sirius, a fusion protein between CBM3a and mRFP1, to prove the CBM3a platform interaction with cellulose. Once produced and purified, we incubated the protein with cellulose and after several washes with resuspension buffer, we measured the fluorescence in cellulose pellets (Figure 1).

As expected, we demonstrated that the CBM3a platform interacts with cellulose since the fluorescence is only present in the Sirius sample.

Then, we designed Orthos, a fusion protein between CBM3a platform and monomeric streptavidin linker, which displays high affinity for biotinylated compounds. We mixed Orthos protein with biotinylated fluorophore (FITC and mtagBFP separately), followed by incubation with cellulose. After several washes with resuspension buffer, we are able to measure fluorescence in cellulose pellets samples (Figure 2).

Fluorescence intensity is twice higher in Orthos-FITC sample than with controls samples (buffer and control FITC alone), and this difference is statistically significant (p-value of 0.1 for Mann Whitney test). Therefore, we demonstrated that our Orthos protein can be conjugated, via streptavidin linker, to biotinylated compounds.

And finally, we designed Cerberus, which is similar construct to Orthos (i.e. containing an amber stop codon at it C-terminal linker). Its production in the presence of the UnAA AzF within the culture medium led to the full length protein consisting in a binding molecular platform to cellulose and two independently functionalizable linkers, including the ready-to-use AzF. In order to validate this latter, we performed a click chemistry reaction between Cerberus and DBCO-bearing fluorescein. We then incubated this sample with cellulose, and after several washes with resuspension buffer, fluorescence intensity can be measured in cellulose pellets (Figure 3).

Fluorescence intensity is 3 times higher in Cerberus-FITC sample than controls samples (buffer and control FITC alone), and this difference is statistically significant (p-value of 0.03 for Mann Whitney test). Thereby, we demonstrated that our Cerberus protein can be conjugated, via its UnAA AzF, to organic or inorganic molecules bearing a DBCO group by click chemistry.

So we proved the binding potential of each head, which permitted us to validate our protein design.

Compounds Fuctionalization

---

Additional to our Cerberus production, we functionalyzed some of our compounds either with by biotinylation (in vivo and in vitro) or by chemical coupling of an alkyne function.

Indeed, for the streptavidin linker, we biotinylated in vivo blue fluorescent protein by designing a protein with an Avitag, recognition site of the biotin ligase BirA. We also chemically biotinylated the fluorescein using the copper-free click chemistry to ligate in vitro a biotin containing a dibenzocyclooctyne (DBCO) group to fluorescein bound to an azide function. We thus obtained biotinylated FITC. The functionalization of these compounds was tested during the cellulose pull down assay and the results were statistically conclusive (we obtained between control sample and functionalysed sample a p-value of 0.1 for fluorescein).

Concerning the AzF linker, we chemically added an alkyne function to graphene and carbon nanotubes. Samples were analysed by thermogravimetric analysis (TGA, Figure 4). The results displays a significant mass loss at 550 °C for functionalized CNT, and this is not observed for control sample. The difference between these samples permitted us to conclude that the CNT functionalization has been a success.

We performed the same observation for functionalized graphene sample that proved that the functionalization has been a success.

Having the capacity to functionalyze several types of compounds is real advantage because we can bind both organic and inorganic molecules of different size to cellulose. In this way, we can add another functionalities and properties to cellulose.

Production and Functionalization of Bacterial Cellulose

---

Based on the optimization of culture conditions of iGEM Imperial 2014 team, Gluconacetobacter hansenii culture allowed us to produce with success cellulose (Figure 5).

Once separated production of bacterial cellulose and Sirius (fusion protein between CBM3a and mRFP1), we incubated both together and, after several washes with resuspension buffer, revealed on UV table (Figure 6).

As we can observe, only the cellulose sample incubated with Sirius (CBM3a-mRFP1) remains fluorescent after several washes with resuspension buffer. The control cellulose sample showed any residual fluorescence after incubation with mRFP1 alone (not coupled with CBM3a platform). These results shown that CBM3a interacts with bacterial cellulose, and allows fixation of the mRFP1 protein to cellulose and give to cellulose new perspectives of functionalization.

Conclusion

---

To conclude on all our experiments, we validated the three binding heads of our Cerberus protein: the Carbohydrate Binding Module (CBM3a) platform, and the Streptavidin and the AZF linkers. We also validated the functionalization of regenerated cellulose. Moreover, we were a success functionalyze bacterial cellulose with Sirius, that is promising to functionalize this last with Cerberus coupled with molecules.