General Considerations of Improvement
With MEtaPhos, we aimed to provide a new recycling method for phosphate. As a matter of fact, effectiveness and economical aspects are crucial for any recycling process. Due to constraints of time and resources, we could not perform all experiments we wanted and had to limit our pursued approaches. But as we believe in the potential of our project, we hope to advance research on phosphate recycling and lay the base for actual application of a modified phosphate binding protein (PBP).
Optimization of the fusion protein
The fusion protein used in MEtaPhos consists of a PBP and an optogenetic photosensitizer or switch to control the PBP’s binding behaviors. Since we have selected the most promising candidates to work in a short amount of time, there is still potential for optimization here.
The PBP itself has been selected because of its good binding capabilities. 2ABH is a monomer, making it difficult to use in combination with an optogenetic switch. As the optogenetic switches cause a change of conformation, a dimeric PBP would be advantageous for our target application. An interesting candidate to use as an optimized PBP is the wild type of phosphate carrier (PIC) from Saccharomyces cerevisiae. The PIC protein is only active in its dimeric form, while the monomeric subunits are shown to be inactive . By presuming this context, a well-functioning switching system could be created in which the subunits of the optogenetic switch are fused with the PIC monomers. Upon illumination, the PIC monomers are brought into contact by the conformational change of the switch and dimerization is initiated. Upon dimerization, the PIC is active until the switch is triggered again and the close contact of the PIC monomers and therefore its dimerization is disrupted. Although the PIC subunits composition is favorable for the use in MEtaPhos, it originally is not a PBP, but rather a transport protein. Its binding capabilities towards phosphate would have to be checked before further use.
Another possibility to create dimeric phosphate binding proteins would be to create split PBPs. By separating a PBPs amino acid chain, an artificial dimer could be created. The created monomers would be inactive and, after triggering the switch, the PBP could be reactivated again. The used PBPs could be optimized using well-documented methods like QuikChange as well. But because these processes are complicated and require a lot of repeated experiments and time, we could not perform experiments following this idea.
The used optogenetic switch could be optimized as well. After we proved our concept (see Results) using the photosensitizer SOPP3, you can see in figure 1 how SOPP3 works.
Figure 1: SOPP3 function 
We worked with VVD as it is the optogenetic switch we could gather the most information on. VVD itself has the important downside that its reversion time in darkness is very long. In figure 2 you can see how VVD works.
Figure 2: VVD function 
This prolongs the time required for phosphate release and thus limits the application possibilities. Nevertheless, the darkness reversion time can be optimized through engineering, as the optogenetics expert Prof. Möglich told us in an interview (see Human Practices). An optimized VVD domain could be very useful in future applications. Prof. Möglich told us about an optogenetic switch which works with blue light as well and already has a much shorter darkness reversion time than VVD. The PtAU1-LOV domain has a darkness reversion time in the range of tens of minutes and may provide further insight into the use of a fast-reversing blue light switch in our project. In figure 3 you can see how PtAU1-LOV works.
Figure 3: PtAU1-LOV function 
Since research on optogenetic switches has intensified in recent years, many different switches covering different wavelengths have been discovered. In addition to switches operating with blue light, many optogenetic switches triggered by red light and far-red light have been published recently. Since red light is less reflected by aqueous media, a red-light switch might be a better choice for MEtaPhos. The exact effectiveness of blue light compared to red light in large volumes of water should be investigated before a final decision is made on the wavelength used. We talked to Prof. Möglich about optogenetic switches which are triggered by red light. One such interesting optogenetic switch is PhyB-PIF6. In figure 4 you can see how PhyB-PIF6 works.
Figure 4: PhyB-PIF6 function 
The most important advantage when compared to VVD is its ability to be switched off by far-red light illumination of 740 nm. It is activated by red-light illumination of 660 nm. Since the excitation and reversion time is in the range of milliseconds, it is a favorable candidate for an optogenetic switch, provided the red-light illumination is as good as the blue-light illumination or even better. The protein of interest used with PhyB-PIF6 must be a heterodimeric protein. If the PBP of interest is a homodimeric protein, Cph1 might be a better candidate for the optogenetic switch than PhyB-PIF6. In figure 5 you can see how Cph1 works.
Figure 5: Cph1 function 
Cph1 has similar properties as the PhyB-PIF6 switch but allows the use of homodimeric proteins of interest. We believe that an optimized optogenetic switch could improve our approach, and therefore want to point out the above-mentioned alternatives.
Experiments with Real or Composite Wastewater
Our focus was to prove our concept with an optogenetic switch, so we used simple phosphate solution as our fake wastewater. With Dr. Mayer, an expert for phosphate binding proteins, we had an interview (see Human Practices), where she pointed out that testing our approach in real or composite wastewater is necessary to find out how the protein behaves with other molecules like organic molecules. One possible step could be using phosphate with yeast extract to simulate how the protein would behave with organic molecules. Another advantage of testing the behavior of the optogenetic switch in real/composite wastewater is that it is possible to test which wavelengths, for example, blue or red, are suitable for the application. It is unknown which wavelengths reach the optogenetic switch the best in real/composite wastewater. On the one hand, blue light is reflected by water, on the other hand, red light is absorbed by water. Both red or blue light have advantages and disadvantages for the optogenetic switch, but to find out which wavelength is suitable, only testing or simulating it can give answers.
Optimization of the Protein Production
MEtaPhos requires a large amount of protein, which is related to the fact that it is to be used in sewage treatment plants (see Implementation) and the associated large amounts of wastewater containing phosphates. Therefore, the production of the fusion protein would be a key part of an industrial realization. First, the use of a cell-free medium for protein production would improve the biosafety of the process by eliminating the need for genetically modified organisms. We have been in contact with the company LenioBio, which has developed a cell-free protein expression platform. Although we have not had enough time to apply this technique ourselves, we assume it could significantly improve protein production. In the best case, protein production could be centralized and automated, which would greatly reduce the costs for protein production and make the process more economical.
Points of Application
After talking to farmers and experts (see Human Practices), we concluded that MEtaPhos could be used in multiple ways. Sewage treatment plants can be a possible use case because they need to regain phosphate in sewage sludge in the EU. Different implementation locations of operations have advantages and disadvantages. If we use MEtaPhos in a decentralized approach, we propose the use of membranes in which the proteins are immobilized and through which the effluent can flow after the final settling tank. Building columns are expensive, and we would deal with a huge mass of water that will constantly flow through the columns.
Another possible implementation location of operation is a centralized approach, where the ash of the mono-incineration plants could be resuspended to recycle phosphate. This has the advantage that we could collect the ash from different sewage treatment plants and only need a few specialized sewage treatment plants to recycle phosphate.
For more information please have a look at the Implementation.