Virus Disinfection from Environmental Water Sources Using Living Engineered Biofilm Materials

Abstract Waterborne viruses frequently cause disease outbreaks and existing strategies to remove such viral pathogens often involve harsh or energy‐consuming water treatment processes. Here, a simple, efficient, and environmentally friendly approach is reported to achieve highly selective disinfection of specific viruses with living engineered biofilm materials. As a proof‐of‐concept, Escherichia coli biofilm matrix protein CsgA was initially genetically fused with the influenza‐virus‐binding peptide (C5). The resultant engineered living biofilms could correspondingly capture virus particles directly from aqueous solutions, disinfecting samples to a level below the limit‐of‐detection for a qPCR‐based detection assay. By exploiting the surface‐adherence properties of biofilms, it is further shown that polypropylene filler materials colonized by the CsgA‐C5 biofilms can be utilized to disinfect river water samples with influenza titers as high as 1 × 107 PFU L−1. Additionally, a suicide gene circuit is designed and applied in the engineered strain that strictly limits the growth of bacterial, therefore providing a viable route to reduce potential risks confronted with the use of genetically modified organisms. The study thus illustrates that engineered biofilms can be harvested for the disinfection of pathogens from environmental water samples in a controlled manner and highlights the unique biology‐only properties of living substances for material applications.

Briefly, a recombinant gene combining CsgA and its biological secretion signal (ss) sequence appended with C-terminal C5 peptide tags was obtained by polymerase chain reaction (PCR The synthetic materials used in this work that constitute the plasmids are described in Supplementary Table S1. The plasmids are described in Supplementary Table S2. The   strains are described in Supplementary Table S3. The primers used for qPCR are described in Supplementary Table S4.
The detailed information for construction of this strain was described in a previous publication from the same group [1] . E. coli MG1655 PRO ΔCsgA generated by removing the kanamycin resistance cassette was described previously [2] (in order to free this antibiotic selection marker for subsequent usage was added to the column to elute target proteins. The protein concentration was detected via a Nanodrop 2000 Spectrophotometer (Thermo). The proteins were further confirmed via SDS-PAGE and western blot, following the protocols described in a previous work [3] . The hemagglutinin protein (5 mg/mL), provided by Shanghai Institute Biological Products Co. Ltd., was loaded onto the silicon-coated substrate at a rate of 2 μL/min followed by washing with TBS buffer (20 mM Tris·HCl, 150 mM NaCl, pH=7.4) at a rate of 10 μL/min, then 2 mg/mL BSA protein solution was loaded at a rate of 2 μL/min to block the substrate and 0.6 mg/mL CsgA-C5 or CsgA protein monomers were loaded onto the substrate at a rate of 2 μL/min followed by washing with TBS at a rate of 10 μL/min. Experiments were performed at 25°C. water. The target proteins were confirmed using SDS-PAGE and western blot; data are shown in Supplementary Fig. 1a.

X-ray fiber diffraction
Mature fibrils formed by different protein samples were pelleted by centrifugation, followed by washing with excess distilled water several times to remove remaining salts. Fibril pellets were then suspended with 5 μL distilled water. Suspensions (2 μL) were pipetted between two fire-polished glass rods followed by drying for a couple of hours. The diffraction data were collected by an in-house X-ray machine equipped with a Rigaku Micromax-007 X-ray generator and an R-Axis IV++ area detector.

Biofilms cultivation:
Seed cultures (Tc Receiver /CsgA-C5) were inoculated from frozen glycerol stocks and grown in LB medium using chloramphenicol antibiotics at 34 μg/mL. Seed cultures were grown for 12 h at 37°C in 14-mL culture tubes (Falcon), with shaking at 220 rpm. Cells collected from the above seed cultures through centrifugation were resuspended with ddH 2 O, and were then added into M63 medium at a volume ratio of To demonstrate that the CsgA-C5 biofilms can capture virus in solution as shown in Figure 3, gradient titers of virus were added into the culture solution directly and co-cultured for 3 days at 29°C before characterization.

Cell density detection and biofilms quantification
After cultivation of biofilms for 3 days at 29°C, the supernatant was removed, and the biofilms were scraped from the bottom of the culture dish and resuspended in 1 mL TEM images were obtained on a FEI T12 transmission electron microscope operated at 120 kV accelerating voltage.

Scanning electron microscopy (SEM) imaging
The

Enzyme linked immunosorbent assay (ELISA)
For the ELISA assay, we followed the general ELISA protocol of Sino Biological. were carried out at Wuhan Institute of Virology following a published protocol [4] .
Infectivity titers for all virus stocks were calculated as plaque forming unit (PFU) per mL.  The cell nuclei were conjugated by Hoechst33342 (1:1000) (Beyotime). The samples were observed with an inverted fluorescence microscope (Olympus IX83).

Homology modelling and molecular docking
We used MODELLER [5] for homology modelling and Schrödinger to set up the initial fiber and docking model. The initial structure of CsgA was from the last frame of 1 μs molecular dynamics from a previous study [3] . The structure of HA protein was obtained from PDB structure (1HGG chain: C, D) with the deletion of original sialic acids. The initial structure of C5 peptide was gathered from the de novo peptide build in Schrodinger Suite. The highest scored docking result to HA protein sialic acid binding domain was chosen as a candidate of C5 peptide. We then built monomer and pentamer fibrillar CsgA-C5 based on the candidate by MODELLER. The best scored candidate was chosen from 50 generated structures of monomer and pentamer fibrillar CsgA-C5. Then we performed an induced fit docking (IFD) on monomer CsgA-C5 to HA protein to get the structure of complex. Docking results were performed by Glide XP [6] and IFD [7] modules in Schrodinger Suite.

Molecular dynamics simulation
We performed all atom molecular dynamics simulations (MD) for 3 systems. cut-off of long-range nonbonding interactions was set at 10 Å. MM-GBSA [9] (Molecular Mechanics/Generalized Born Surface Area) score from Schrodinger Suite was used for analyzing MD results to get binding free energy of CsgA-C5 to HA protein during molecular dynamics. Visual Molecular Dynamics (VMD) [10] was used to capture 8 stable frame during 800ns MD of system (1)

Hemagglutination inhibition assay
Virus at titers of 7×10 4 PFU/mL was seeded onto a 12 well plate in 3 mL M63 culture medium with and without addition of Tc Receiver /CsgA-C5 and incubated at 29°C for 72 hours. A total of 100 μL supernatant from the culture medium was removed and added to a V-bottom well then mixed with 100 μL 1% chicken red blood cells in PBS buffer. After incubation for 1 hour at 37°C, the plates were then photographed. Virus disinfection from river water using CsgA-C5 biofilm-coated industrial fillers.

RNA extraction, reverse transcription, and qPCR
Five pieces of CsgA-C5 biofilm-coated industrial fillers were put in a column tube and incubated with 10 mL influenza virus solution in river water at room temperature for at least 2 hours. The solution was then allowed to flow out at an average rate of In order to quantify the biomass of CsgA-C5 biofilm coated on fillers, five pieces of fillers coated with or without CsgA-C5 biofilms were immersed into 5 mL Congo red solution (0.04 mg/mL) and incubated for 1 hour. The amount of Congo red bound by CsgA-C5 biofilms were quantified by subtracting the A 496 nm of biofilm-coated fillers