Figure 1: Protein extraction kit -- specific antibodies are fused to the C terminal of mamC protein on a magnetosome membrane

One of the first application with our magnetosome and insertion kit is protein extraction (Figure 1). Protein extraction (also named protein purification) is a series of processes intended to isolate one or a few proteins of interest from a complex mixture[1].

Among a bunch of extraction/ purification techniques, affinity ligand techniques represent currently the most powerful tool available to the downstream processing both in term of their selectivity and recovery. Though standard liquid column chromatography is currently the most often used technique for the isolation and purification of target proteins and peptides, one of the biggest disadvantage of it is that it is not capable to cope with samples containing particulate material. This make early stages of purification process, where suspended solid and fouling component are found in sample, unmanageable. Besides, magnetic separation is usually comparatively gentle to the target proteins or peptides. This avoids the problem of having larger protein complex to be broken up by traditional column chromatography techniques. [2]

Using magnetic separation is great, however why use our magnetosome but not just chemically produced magnetic beads? And here are the reasons:

Figure 2: Lower efficiency of binding due to non-specific orientation of antibodies by chemical methods

The magnetosome we produce has a smaller size (30nm-120nm) than traditional magnetic beads (1-4.5µm), so magnetosome with antibodies could have a higher binding efficiency due to the bigger surface area-volume ratio. [3] Besides, using our biological construct design, we are able to fuse any protein of interest to the transmembrane protein of magnetosome(mamC)[4], so we can add some antibodies on the membrane of magnetosome and we can isolate the antibody-antigen complex by magnetic force. Chemical methods currently can attach antibody to synthesized magnetic beads, however, the orientation attachment of it is not a specific one. This not only cause a large decrease in the availability of binding sites but also a easier detachment of the antibody from the magnetosome compared to our biological method (Figure 2).

Figure 3: The binding of GFP to GFP nanobody on membrane of magnetosome

To test our application, we chose to use GFP protein as our target protein as it is easy to detect and thus be characterized (Figure 3). To target the GFP protein, we will first need to fuse the green fluorescent protein (GFP) nanobody on the membrane of magnetosome. This requires the dissociation constant of GFP: GFP-nanobody complex is 1.4x10-9 M [5], which indicates a high affinity of binding, so we can easily extract the GFP and GFP-tagged protein by magnetic force.

Water treatment

Figure 4: Heavy metal binding peptide are fused to mamC proteins on the magnetosome membrane. The magnetosomes are captured by magnetic bars while water with heavy metals flows through.

Heavy metal is one of the major components in marine pollution due to industrial wastes, vehicle emissions, lead-acid batteries, aging water supply infrastructure and more. Different kinds of heavy metal ions, such as Pb, Cu and Ni, are found in the polluted water system[6]. Moreover, owing to the lead-in-water scandal happened recently in Hong Kong[7], we decided to use our engineered magnetite help tackle the problem as well. The principle of doing so is more or less the same as that of protein extraction. However, instead of fusing an antibody behind the trans-membrane protein mamC[4], metal binding peptide will be fused behind instead. The gene sequences for the specific metal binding peptides of different metals are actually known through previous researches. By expressing different heavy metal binding proteins onto magnetic beads, heavy metal ions could then be captured and be easily removed from water by magnet. The heavy metal can then be released from our binding peptide using EDTA solution so that our magnetosome can be reused again (Figure 4). To test the efficiency of our application, we decided to test the Pb-binding peptide this time. The gene encoding the lead binding peptide sequence TNTLSNN and the maximum adsorption loading (qmax) of lead is 526 μmol/g dry cell weight[8].

Characterization of lead binding peptide (LBP) efficiency

Lead binding peptide (LBP) TNTLSNN was designed to bind with Ni-ATA by adding His-tag (HHHHHH). We generated four set of peptides:

1) 1xLBP-6xHis TNTLSNNHHHHHH

2) 2xLBP-6xHis TNTLSNNTNTLSNNHHHHHH

3) 2xLBP-linker-6xHis TNTLSNNGGGTNTLSNNHHHHHH

4) 1xLBP-linker-6xHis TNTLSNNGGGHHHHHH

to investigate whether 1. number of LBP; 2. linker between lead binding site and his-tag site affect the lead binding efficiency. Final concentration of 1 mM lead nitrate solution mixed with 1mg of each peptide in total 1 mL was incubated for 1 h at RT. Negative control: 1. peptide without lead nitrate solution; 2. lead nitrate solution only; 3. buffer only was setup to assess the lead binding effect of the peptides. 200uL Ni-ATA resin (cOmpleteTM His-Tag purification resin) was used to capture the peptides. The resin was washed three times with Ni-ATA buffer (50 mM NaH2PO4, 300 mM NaCl, pH 8.0), and eluted with 1ml Ni-ATA buffer plus 100 mM imidazole. 150 uL Elution was mixed with 450 uL concentrated nitric acid, incubated for 24 h at 60 C, and loaded into Atomic Absorption Spectrometer (AAS). Lead nitrate solution standards were prepared to calculate the lead concentration in samples.

Compared to the negative control 1. peptides only and 3. buffer only, higher Pb concentration of elution was found in 2. Pb only control. It indicated there is either non-specific binding of Pb on the Ni-ATA resin or incomplete washing. Compared peptide+Pb samples with 2. Pb only control, 1xLBP-6xHis with or without linker showed higher concentration of Pb binding than that with 2xLBP, suggesting 2xLBP may hinder the 3D-configuration for Pb binding.

Microbial Fuel Cell

The design of our microbial fuel cell involves a simple construct inside the bacteria Azotobacter vinelandii which consist of hydrogenase and OprF porin proteins. The expression of the HoxKGZ genes would produce membrane bound hydrogenase which converts hydrogen to protons and electrons inside the bacteria (H2 -> 2H+ + 2e-)[9].

The electron produced will then transport along the periplasmic space and be picked up by the electron acceptor. After the binding of an electron to the electron acceptor, the reduced electron acceptor will be transported out of the cell through the OprF porin [10] and possibly other membrane channels to the anode chamber. With OprF porins added to the bacteria, the reduced electron acceptors can be brought out of the cell to the anode chamber more effectively. And if possible, with mutated OprF porins which have a higher open probability added to the cell membrane of the bacteria[11], it could further increase the efficiency of electron acceptors traveling in and out of the cell. In other words, more electrons could reach the anode chamber at a higher rate.

Figure 5: Model of our microbial fuel cell

After the reduced electron carriers escape the bacterial cell, the electrons will then stick to the anode and this release of an electron will free the electron acceptor. The oxidized electron acceptor will be transported back to the periplasmic space of the Azotobacter vinelandii to pick up another electron and so on. The electrons then pass through the circuit to the cathode, and reduce the oxidant (oxygen) in the cathode chamber. Through the continuous reduction and oxidation processes, an electric current will be generated for our microbial fuel cell.

Furthermore, this is also where our magnetosome come into play. Together with our magnetosome expressed in the microbial fuel cell bacteria, Azotobacter vinelandii. The bacteria can be brought to the electrode with much closer physical contact rather than randomly dispersed within the culture solution. With a shorter diffusion distance, the diffusion rate for the electron to the electrode can be greatly increase. Ultimately, the apply of magnetosome can contribute in the improvement of efficiency of existing microbial fuel cells.

Reference

1. AHMED, Hafiz. Principles and reactions of protein extraction, purification, and characterization. CRC Press, 2004.

2. SAFARIK, Ivo; SAFARIKOVA, Mirka. Magnetic techniques for the isolation and purification of proteins and peptides. BioMagnetic Research and Technology, 2004, 2.1: 7.

3. THURBER, Greg M.; WITTRUP, K. Dane. A mechanistic compartmental model for total antibody uptake in tumors. Journal of theoretical biology, 2012, 314: 57-68.

4. XU, Jun, et al. Surface expression of protein A on magnetosomes and capture of pathogenic bacteria by magnetosome/antibody complexes. Frontiers in microbiology, 2014, 5.

5. KUBALA, Marta H., et al. Structural and thermodynamic analysis of the GFP: GFP‐nanobody complex. Protein science, 2010, 19.12: 2389-2401.

6. JÄRUP, Lars. Hazards of heavy metal contamination. British medical bulletin, 2003, 68.1: 167-182.

7. Gloria Chan. Lead 80 times the safe limit found in water at Hong Kong public housing estate where scandal broke. South China Morning Post, 2015.

8. NGUYEN, Thuong TL, et al. Selective lead adsorption by recombinant Escherichia coli displaying a lead-binding peptide. Applied biochemistry and biotechnology, 2013, 169.4: 1188-1196.

9. SAYAVEDRA-SOTO, LUIS A.; ARP, DANIEL J. The hoxZ gene of the Azotobacter vinelandii hydrogenase operon is required for activation of hydrogenase. Journal of bacteriology, 1992, 174.16: 5295-5301.

10. YONG, Yang‐Chun, et al. Enhancement of extracellular electron transfer and bioelectricity output by synthetic porin. Biotechnology and bioengineering, 2013, 110.2: 408-416.

11.SUGAWARA, Etsuko; NAGANO, Keiji; NIKAIDO, Hiroshi. Factors affecting the folding of Pseudomonas aeruginosa OprF porin into the one-domain open conformer. MBio, 2010, 1.4: e00228-10.