Column

Phosphix Column

Phosphix is proposing an innovative water treatment method designed to capture and recycle phosphate from wastewater through a strategic and effective approach. Inorganic phosphate (Pi) is required for many processes, but the primary demand lies in fertiliser use. The current depletion of phosphate mineral reserves has caused a spike in the price of Pi, with access to the resource being politically and economically controlled. Only 10-25% of phosphate is uptaken from fertiliser, making this an inefficient process as well as causing leaching into the waterways (Johnston et al, 2014). Phosphate within water systems poses an environmental threat, causing eutrophication and algal blooms, therefore recapturing phosphate would not only enable the generation of a cyclic phosphate economy, but it also decreases the environmental harm caused by fertiliser usage. While many regions already have established water treatment systems, enhancing these existing infrastructures to recover phosphate proves more practical and cost-effective than developing entirely new ones.

PafA converts organic phosphate into Pi, the form accessible to plants. This means that if coupled to a phosphate capturing system, all phosphate captured could be converted to an accessible form, increasing the yield of capturing methods to hopefully make it more economically feasible to resell captured phosphate for use in fertiliser. As seen by Dr Cassidy from Nibeenabe, when the financial barrier to implementing a more sustainable approach is removed, even disadvantaged areas - populations often more heavily impacted by the problem - can alleviate the environmental harm. Therefore it was important to us in our prototyping to think of a scalable method that could also be implemented in more remote environments.

Current Alternatives to Mined Phosphate Fertilisers

Current alternatives to mined Pi, such as calcium phosphate (CaPO₄) or struvite (MgNH₄PO₄•6H₂O) precipitation, serves to recover phosphates from sewage. However, this can generate fertilisers containing potentially harmful substances such as heavy metals or micropollutants (such as aspirin, paracetamol, and caffeine) due to the use of human waste and the sewage treatment process. Furthermore, struvite production can also lead to nitrous gases being produced. Previous research has isolated calcium phosphates from household wastewater (containing urine and faeces) - despite the success in recovering phosphates, fertiliser produced from blackwater is not allowed to be used for food production within European regulations as the water contains human waste. Sewage has been an attractive source of phosphate recovery as it is produced in high volume and contains high levels of phosphate. However, most sewage collection and processing methods lead to contamination with high concentrations of heavy metals, so this cannot be implemented as a viable source of phosphate. Further problems with this method included the high concentration of solids in the reactor preventing ideal conditions from being met (Cunha et al., 2020). There is also a high chemical cost to this method, as chemicals such as ferric chloride must be used.

Bio-struvite can be produced via bacteria such as Brevibacterium antiquum from wastewater. It co-precipitates with significantly less heavy metals, allowing the recovery of as much as 97% of phosphates depending on the initial phosphate concentration and bacteria used. This version of struvite met the EU regulations and has the potential to be used as an inorganic fertiliser. However, this methodology is limited by logistics: design of reactors, growth conditions, and competition of microorganisms in reactors as the microbe-driven reactions are generally slower. In order to increase the speed, the conditions must be optimised, meaning the reactors must be complex to ensure constant conditions (Leng and Soares, 2023). Additionally, the exact process by which biomineralization occurs is not entirely known. It has been theorised that it may occur naturally as a byproduct of metabolism or extracellular substances could cause the formation of the crystals (Arias, Cisternas and Rivas, 2017), but a lack of insight into these molecular mechanisms hinders optimisation efforts.

Biomineralization involving anammox reactors to produce crystallised phosphate has also been attempted. In phosphate-enriched mediums, the yield is around 5-7% and is obtained through granules containing phosphates. Minerals are collected as they sink to the bottom of the reactors as they gain mass so are easily collected (Magrí et al., 2020).

Another method that has been explored is the use of phosphate-binding proteins from E.coli (Venkiteshwaran, Wells and Mayer, 2020). These have advantages over other methods as ion exchange methods can’t distinguish between arsenate and Pi due to their similar chemical structure, and this causes issues in the potential for downstream use due to the toxicity of arsenate. This method generates a higher purity, with 97% of the recovered material being Pi, see figure 1. PBP was immobilised onto NHS-activated Sepharose beads (Venkiteshwaran, Wells and Mayer, 2018). High pH (~12.5) causes the desorption of Pi from PBP, but does not damage the folding of PBP, enabling it to be reused. The disadvantage of this method is that these wash steps complicate the extraction, and might make it more expensive. The key issue with the use of proteins for reclamation from wastewater is deciding upon the immobilisation technique that will allow us to continually reuse the proteins to make it an economically viable option (Hutchison et al, 2021)

Figure 1. taken from Venkiteshwaran, Wells and Mayer (2020): Exposure of PBP to either arsenate or Pi. Adsorption was carried out at pH7 at 25C, and desorption at pH 12.5, 25C. Error bars represent SD from three repeats.

Aachen 2022 team engineered a phosphate capture system with light-activated phosphate release. They achieved this by fusing a phosphate-binding protein (2ABH- BBa_K4138000) to a photosensitizer (SOPP3) so it released its cargo in response to light instead of high pH levels. From figure 2, 2ABH decreases the concentration of phosphate in the stock from around 290uM to ~200uM, capturing a small amount of Pi. This coupled with the photosensitizer SOPP3 increased phosphate binding , so around 200uM of phosphate was removed from solution (figure 2).

Figure 2: The impact of illumination treatment applied by Aachen 2022 on two proteins: blue= SOPP3-2ABH and orange- 2ABH. Differences in concentrations of phosphate binding are likely due to mismeasurements in the amount of protein that was immobilised rather than the SOPP3 improving the binding efficiency of 2ABH.

Whilst current methods using precipitation are able to remove large amounts of phosphate, the chemical waste produced and more supplies required than a reusable enzyme system meant we were enticed by the work of Aachen 2022. For our prototype, we didn’t have the facilities to use the illumination treatment, and this may not be feasible for all wastewater treatment facilities, so we planned to use the 2ABH without the photosensitizers as it is capable of releasing phosphate under high pH conditions. This does come with more chemical waste, so in more remote areas, whilst the initial costs to purchase the LED systems needed for the illumination treatment would be higher, it may reduce costs from transport of chemicals to a remote location to use SOPP3 coupled 2ABH.

Immobilising PafA and 2ABH

In our research, we explored various methods of using enzymes in water treatment. Free enzymes, though highly effective, tend to lose activity quickly and cannot be reused, making them costly and inefficient for large-scale applications like wastewater treatment. On the other hand, immobilised enzymes, which are attached to or combined with various matrices, offer enhanced stability and reusability, significantly reducing overall costs and waste. Enzyme mediators, acting as electron shuttles, can further improve enzyme activity and pollutant degradation, though their effectiveness depends on specific conditions.

To apply this knowledge to water treatment, we will focus on enzyme immobilisation to prevent enzyme loss. Our initial plan was to immobilise the enzyme onto glass beads to bind the enzymes, as suggested to us by Dr Routoula. Glass beads are already used in wastewater treatment, making it a trialled and tested method. Ideally, we would couple the enzyme to the glass bead using very stable covalent bonding; given the scope of the iGEM timeline and the use of dangerous chemicals such as Piranha solution, it wasn’t possible of us to immobilise our PafA. Dylan Lewis showed us how to immobilise using the His-tag on our proteins that have been purified using IMAC, however we weren’t able to repeat this process for our enzyme due to the dangers of Piranha solution.By encapsulating the enzymes on beads, we allow substrate flow while keeping the enzymes in place. This bead-based immobilisation approach will serve as the foundation of our method for phosphate recovery in wastewater treatment within a column device to enable the flow of wastewater through the beads. For safety purposes and to check if our enzyme would still function when immobilised, we planned to immobilise them on sepharose beads, utilising the method of purifying them to stabilise them. This is very limited and couldn’t be used for within wastewater treatment as the sepharose would be degraded by microbes etc, and the link between Nickel and His tag is not strong enough to withstand the harsh conditions of wastewater.

To test the practicality of the column device we conducted several experiments. We started by preparing 4 different columns as depicted in figure 3.

Figure 3: PNPP put into 4 columns containing different enzyme mixtures. The colour of the column depicts the ideal results of this prototyping step. The PBP we used is 2ABH.

We utilised the PBP (2ABH) as a method of binding the inorganic phosphate released by the hydrolysis of PNPP within the experiments. This allowed us to be able to activate and deactivate the PBP to release the bound inorganic phosphate when necessary; this would be critical for the collection of the inorganic phosphate from the wastewater flowing through the column.

We began by focussing on columns 1 and 2:

Column 1 (Control of no enzyme):

Take an empty P10 column, and fill with 5ml of the Sepharose Bead slurry - this 5ml is what we'll refer to hereafter as a column volume (CV).
When not in use, keep the lid and the tip of the P10 column on and store the bead slurry in 100% ethanol.
When ready for use take the following steps to wash and "charge" the column
Wash through with ~4 CV of MQ H₂O
Charge (i.e., add the Ni (II) ions) the column by adding ~4 CV of NiSO₄ - this should stain the beads a teal colour
Wash through again with water as before
Wash through ~4 CV of binding buffer to equilibrate the column
Place the cap and lid of the column back on when not in use
Wash through with binding buffer ~4CV

Column 2 (PafA Only):

Add PafA (His-tagged phosphatase).

Washing and Preparation:

Wash with ~4 CV of MQ H₂O.
Charge with ~4 CV of NiSO₄, turning the beads teal.
Wash with ~4 CV of MQ H₂O.
Equilibrate with ~4 CV of binding buffer.
Add PafA to the column to assess its binding to the nickel-charged matrix.
Wash through with binding buffer ~4CV.

Columns 1 and 2 then had 5ml of 40mM PNPP solution flown through and the following samples were collected and absorbance measured at 405nM.

Samples

Sample	Absorbance at 405nm
Blank (PNPP sol)	-
PafA Flowthrough	-0.882
Control	-0.962
PafA 1 (1st column)	0.669
PafA 2 (2nd column, 1st round)	-1.334
PafA 3 (2nd; 2nd round)	-0.852
PafA 4 (2nd; 30 min batch)	1.202

Samples were taken for:

PNPP mother solution (40mM)
PafA flowthrough - the flowthrough during the process of attaching PafA to the beads
Control - the flowthrough of PNPP solution through column without enzyme
PafA 1 - the first flowthrough for PNPP 40mM through first column with PafA
PafA 2 - the first flowthrough for PNPP 40mM through second column with PafA
PafA 3 - the second flowthrough for PNPP 40mM through second column with PafA
PafA 4 - the third flowthrough for PNPP 40mM through second column with PafA, the process is considered as batch (the solution was left in the column for ~20 minutes before collection)

The absorbance values indicated the need for us to conduct batch processes when utilising the column device, as these values were significantly higher than the absorbance values of flow through. As a result of this we conducted the experiment again with batch processes implemented; and gathered the following results:

Sample	Absorbance 405nm	After 20 min +2.2 batch	After another 20 min +2.3 batch
Blank	-	-	-
1.1	-0.189	0.277	-
2.1	0.034	0.459	0.062
3.1	0.261	0.612	0.243
4.1	0.416	0.474	0.425
5.1	0.286	0.267	0.283
1.2 Batch	0.013	0.521	0.112
2.2 Batch	-	0.184	0.340
2.3 Batch	-	-	0.392

To further prove the concept we run the experiments comparing the effectiveness of nickel columns with just PafA and PafA and Phosphatase Binding Protein (PBP).

Phosphatase Binding Protein (PBP) is a tool used to enhance the binding efficiency of enzymes like PafA to solid supports in biochemical assays. In our experiments, we co-immobilized His-tagged PafA and PBP on nickel-coated Sepharose beads to investigate if PBP could improve enzyme immobilisation and activity. This optimization is crucial for developing more effective phosphate recovery systems in water treatment applications. We created those columns using the following protocol:

Column 3 (Pafa + PBP):

Add both Pafa (His-tagged) and PBP.

Washing and preparation:

Wash with ~4 CV of MQ H₂O.
Charge with ~4 CV of NiSO₄, turning the beads teal.
Wash with ~4 CV of MQ H₂O.
Equilibrate with ~4 CV of binding buffer.
Add a mixture of Pafa and PBP to examine their interaction and potential co-binding to the nickel column.

Column 5 (PBP Only):

Add PBP (assuming no His-tag).

Washing and preparation:

Wash with ~4 CV of MQ H₂O.
Charge with ~4 CV of NiSO₄, turning the beads teal.
Wash with ~4 CV of MQ H₂O.
Equilibrate with ~4 CV of binding buffer.
Add PBP alone to evaluate whether it binds to the column, either directly or non-specifically.

We gathered the following data which compared the use of columns 2 and 3:

Samples	Absorbance at 405nm
Column PafA	-0.710
Column_PafA_PBP	-0.577
Column_PAfA_2	-0.045
Column_PafA_PBP_2	-0.107
Column_PafA_3	0.046
Column_PafA_PBP_3	0.089

We then conducted an experiment to compare columns 2-4 and gathered the following results:

Samples	Absorbance at 405nm
Column_PafA	-0.580
Column_PAfA_2	-0.340
Column_PafA_3	0.006
Column_PafA_PBP	-0.864
Column_PafA_PBP_2	-0.057
Column_PafA_PBP_3	0.023
Column_PBP	-0.834
Column_PBP_2	-0.276
Column_PBP_3	-0.039

Problems that arose with this trial

Unfortunately, PNPP converts to PNP, and we saw that rather than any converted PNP washing through, it seemed to collect in the column, causing the column to become a green colour (See image and video below). This is likely as the PNP was interacting with the positive Nickel ions charging the sepharose beads. As we weren’t able to observe any notable absorbance changes, we didn’t trial column 4 (PafA and the PBP under high pH to release Pi).

The results reported above also may have been impacted by nickel sulphate (a blue solution) or water washing through the column. Therefore, a test for Pi concentration not reliant on measuring absorbance would be more suitable for assaying the effectiveness of this prototype. Future work would be needed to ensure that both the PafA and 2ABH could remain active whilst immobilised and identify a strong form of binding. We can’t take any definitive answers from the data we collected, however going forward these are our plans:

Produce columns in figure 3, however immobilise the enzymes on glass beads using a covalent bond.
Use a stock solution of organic phosphate (such as G6P) of a known concentration and run this through the columns, measuring the amount of Pi present before and after exposure to the enzymes described in figure 3. This could be done using a kit such as EnzChek™ Phosphate Assay Kit by Invitrogen.
Utilising soil samples provided by allotments as a replacement for the use of wastewater in our prototype for the safety of our team. Dissolving this in water, filtering and sterilising it to produce a solution that could go through the column and measuring the amount of Pi before and after putting it through the columns in figure 3. This would allow us to ensure the enzymes could work in harsh conditions for the enzymes.

References

Johnston, A.E. et al. (2014) ‘Chapter Five - Phosphorus: Its Efficient Use in Agriculture’, in D.L. Sparks (ed.) Advances in Agronomy. Academic Press, pp. 177–228. Available at: https://doi.org/10.1016/B978-0-12-420225-2.00005-4.