Our team is approaching this complex problem by means of synthetic biology using natural products produced by microorganisms to design the Enzymatic Flagellulose, a molecular filtering machine.
Our filter consists of a surface made up of cellulose (produced by Acetobacter xylinum) to which bacterial flagella will be immobilized via a cellulose-binding domain. In this case, cellulose itself does not provide any functionality besides acting as a biocompatible carrier. Additionally, it is easy to grow and is a non-plastic material. The single flagellum subunits, also known as flagellin, will be interlinked with plastic degrading enzymes.
Using flagella as a scaffold for enzymes has two major advantages. Firstly, this system enables the creation of a three-dimensional reactive nanostructure that has an increased specific surface with highly catalytic activity. Secondly, flagella may be constructed consisting of various different active sites, which will enable the combination of multiple enzymatic steps in close proximity.



The Enzymatic Flagellulose Flagellins from structure to function:
Although flagella are an assembly of multiple proteins, the flagellum filament is mostly made of one single protein called flagellin. Flagellin proteins consist of four domains (D0-D3). While the D0 and D1 domains are well conserved, the D2 and D3 domains are highly variable in sequence and length (1: Beatson SA, Minamino T, Pallen MJ. Variation in bacterial flagellins: from sequence to structure. Trends Microbiol. 2006 Apr;14(4):151-5. Epub 2006 Mar 15. PubMed PMID: 16540320).
This allows us to modify the D3 domain for our purposes by making use of Synthetic Biology. We designed a flagellin gene (fliC_MCS) with gBLOCKS and introduced specific codons at certain regions of the D3 domain that are supposed to code for the tRNA that will deliver L-Homopropargylglycine (HPG; a non-canonical amino acid). When this gene is introduced into a methionine auxotrophic strain, the bacterium builds in HPG at the certain region of the D3 domain instead of a methionine due to the reason that we do not add methionine to the medium, and the bacterium basically is forced to use HPG. Under mild conditions HPG then reacts with the Azide group in the residue of the noncanonical amino acid Azidohomoalanine that we will introduce specifically to our enzymes.
Below, you will find the detailed reaction of the degradation of PET by the cutinase which has been worked out by the iGEM Team University of California in 2012:

Fig. 5: Degradation of PET by cutinase 2012.igem.org/Team:UC_Davis

As shown here, PET is decomposed to a component which can be transformed into an intermediate of the citrate cycle and terephthalic acid which’s degradation is still too little known.

The Cutinase is an α-β enzyme, which has the typical triad structure Ser 120, His 188 and Asp 175. So the active site is composed of three amino acids. The active serine is not buried under surface loops, so it is accessible to solvents. It is composed of a central β-sheet of five parallel strands covered by five helices on either side of the sheet (Martinez et al., 1992, Fusarium solani cutinase is a lipolytic enzyme with a catalytic serine accessible to solvent). Cutinases are able to hydrolyze the ester-linkage within certain plastics, as Polyethlyeneterephtalate (PET) (C10H804)n.
It could be shown that esterases have a stable pH range from 4,0-8,0, so from acidic to neutral pHs (Faiz et al., 2007, Determination and Characterization of Thermostable Esterolytic Activity from a Novel Thermophilic Bacterium Anoxybacillus genensis A4). Cutinase is an enzyme that once brought to the periplasmic space a leakage occurs that helps to secrete into the extracellular matrix. There are several methods to identify the activity of cutinases. One is to incubate small parts of PET within the enzyme, afterwards an analysis by scanning electron microscopy shows the rising porosity of the plastic. Another way is the p-nitrophenol butyrate (pNPB) assay. pNPB, which is a monomer similar to PET, is degraded. Thereby the absorbance at 405 nm increases. There are different cutinases available for our purpose. Cutinase Hic from Humicola insolens has a temperature optimum at 80°C, which is too high. The cutinase FsC from Fusarium solani ist toxic to E.coli, since our expression is performed in E.coli this cutinase is overruled as well. We decided to use the LC cutinase 2012, since it works at 25°C as good as at 60°C.

The following animation visualizes the concept of our modular filter unit