Biology Cycles of Design

We worked with the iGEM engineering cycle to produce two cycles of design for the biology of our project which are detailed below. Our engineering cycle structure:

Cycle 1 - Cell Free/In vitro activation

Design

Initially, our aim was to extract and activate NosZ in vitro. This cell free system seemed to pose fewer complications in a healthcare setting as no live bacteria are present to potentially infect surrounding organisms, hence biocontainment is less of an issue.

Overview

Our design involved introducing a tagged version of the nosZ gene from P. stutzeri ZoBell into a P. stutzeri chassis in order to make use of the other nos proteins already produced P. stutzeri. Following this, we would extract, purify and activate the tagged proteins in vitro. Finally, immobilisation and testing of the enzymes can occur provided that an appropriate electron donor is supplied.

Copper Site Forms

N2OR has two different copper sites - CuA and CuZ. The CuA site is the point at which electrons enter, whilst the CuZ site is the site of catalysis. The reductase forms a dimer with opposing catalytic sites facing each other. (note: Zhang, L., Wüst, A., Prasser, B., Müller, C. and Einsle, O. (2019). Functional assembly of nitrous oxide reductase provides insights into copper site maturation. Proceedings of the National Academy of Sciences, 116(26), pp.12822–12827. doi:https://doi.org/10.1073/pnas.1903819116. )

The CuZ site can exist in two forms - either CuZ(4Cu2S), or CuZ*(4Cu1S). The difference is the presence (or absence) of a second sulphur atom in the catalytic site.

Both of these forms then have both an oxidised and a reduced state. For CuZ, these are [2Cu²⁺-2Cu¹⁺] and [1Cu²⁺-3Cu¹⁺] respectively. Whilst the reduced state is catalytically active, the turnover number is low, so we instead aim to obtain the CuZ* form. CuZ* also has two oxidation states, which are [1Cu²⁺-3Cu¹⁺] (the resting oxidised state) and [4Cu¹⁺] (the reduced state). Of these, only the reduced state is catalytically active. (note: Carreira, C., Pauleta, S.R. and Moura, I. (2017). The catalytic cycle of nitrous oxide reductase - The enzyme that catalyzes the last step of denitrification. Journal of Inorganic Biochemistry, [online] 177, pp.423–434. doi:https://doi.org/10.1016/j.jinorgbio.2017.09.007. ) Figure 1 Hence, our aim was to obtain extracted enzymes with the CuZ* site, as they demonstrated catalytic activity. Once in the activated state, the enzyme can perform one catalytic cycle. (note: Brown, K.N., Djinović-Carugo, K., Tuomas Haltia, Inês Cabrito, Matti Saraste, Moura, G., Moura, I., Tegoni, M. and Cambillau, C. (2000). Revisiting the Catalytic CuZ Cluster of Nitrous Oxide (N2O) Reductase. Journal of Biological Chemistry, [online] 275(52), pp.41133–41136. doi:https://doi.org/10.1074/jbc.m008617200. )

Extraction

According to Carreira et al., enzyme samples with up to 95% of copper sites in the CuZ* form can be extracted. In order for the enzyme to be extracted in the correct form, three primary methods of lysing bacteria were mentioned:

1. Purification from long term frozen extracts

2. Culturing at pH 6.5 (note: Carreira, C., Nunes, R.F., Mestre, O., Moura, I. and Pauleta, S.R. (2020). The effect of pH on Marinobacter hydrocarbonoclasticus denitrification pathway and nitrous oxide reductase. JBIC Journal of Biological Inorganic Chemistry, [online] 25(7), pp.927–940. doi:https://doi.org/10.1007/s00775-020-01812-0. )

3. Specific anaerobic extraction from one mutant

Method 1 was rejected quickly as it would not be possible for us to freeze extracts for long periods of time. We also decided culturing in lower pHs would be the best option for us due to time constraints. At pH 6.5 and below, NosZ activity seems to get reduced, but the NosZ CuZ* centre is made at higher proportions in more acidic conditions. Hence, culturing at around 6.4-6.5pH would be most beneficial.

Tagging of enzymes

Before lysing bacteria for extraction, we had to find a way of tagging NosZ, so we could isolate it from the lysate and activate only the enzyme required in our project. Two options for tags arose:

1. His-tag
   A tag that typically consists of at least six histidine residues, often at the N- or C-terminus of the protein. Carson et al. showed that the tags generally had no significant effect on the structure of the attached protein, so this tag is safe to use on enzymes. (note: Carson, M., Johnson, D.H., McDonald, H., Brouillette, C. and DeLucas, L.J. (2007). His-tag impact on structure. Acta Crystallographica Section D Biological Crystallography, 63(3), pp.295–301. doi:https://doi.org/10.1107/s0907444906052024. ) Moreover, Liu et al. showed His-tags can be used for Nos genes. Hence, a His-tag gene can be inserted into the bacteria at the end of the gene coding for NosZ, such that tagged proteins are made. (note: Liu, X., Gao, C., Zhang, A., Jin, P., Wang, L. and Feng, L. (2008). Thenosgene cluster from gram-positive bacteriumGeobacillus thermodenitrificansNG80-2 and functional characterization of the recombinant NosZ. FEMS Microbiology Letters, [online] 289(1), pp.46–52. doi:https://doi.org/10.1111/j.1574-6968.2008.01362.x. )

2. Strep II Tag
   Strep II is polypeptide consisting of 8 amino acids that is small and inert, meaning its addition to most proteins causes little to no alteration to the protein’s function, making Strep II an excellent tag. It is attached to proteins by inserting a Strep II sequence on the end of the CDS for the protein. These proteins can then be extracted with the use of a surface containing Strep Tactin, which strongly binds to any proteins with Strep II attached and separates them from any impurities.

We decided on using Strep II tags, as they have been previously used with NosZ specifically. (note: Zhang, L., Wüst, A., Prasser, B., Müller, C. and Einsle, O. (2019). Functional assembly of nitrous oxide reductase provides insights into copper site maturation. Proceedings of the National Academy of Sciences, 116(26), pp.12822–12827. doi:https://doi.org/10.1073/pnas.1903819116. ) Moreover, purified protein obtained by His-tags may need further purification like ion exchange or size exclusion chromatography. Hence, we concluded Strep II was more likely to be successful in our cell free system.

Activation - Electron Donors

Whilst in vivo the electrons for N2O reduction are typically supplied by cytochrome c, electron donors are used in vitro to mimic this process artificially. In order to ensure the CuZ and CuA sites of the extracted enzyme are reduced, we would need to pick an appropriate electron donor for the in vitro approach.

We considered three electron donors in particular, evaluating them based on their cost, reduction potential, solubility in water and safety. Figure 2

Consequently, it was decided that sodium ascorbate would be the most appropriate electron donor not only for the fact that it is the least dangerous and most cost effective, but also as it has proven to be more “physiologically relevant" than methyl viologen in a study focused on the involvement of CuZ in the catalytic cycle of N2OR. (note: Em Biotecnologia, M., Doutora, O., Andrade, I., Galhardas, M., Sofia, D., Pauleta, R., Doutor, A., João, J., De Moura, G., Doutor, S., Dell'acqua, Doutora, M., Maria, P., Correia, Romão, S., Doutora, M., Alice, S., Pereira, Doutora, I. and Andrade, M. (n.d.). Cíntia Catarina Sousa Carreira Insights into the structure and reactivity of the catalytic site of nitrous oxide reductase Outubro 2017. [online] Available at: https://run.unl.pt/bitstream/10362/27900/1/Carreira_2017.pdf [Accessed 26 Sep. 2024]. )

Immobilization

After activation of the enzyme, we investigated the idea of immobilising NosZ in alginate beads. We planned on making alginate beads by ionotropic gelation. (note: Gadziński P, Froelich A, Jadach B, Wojtyłko M, Tatarek A, Białek A, Krysztofiak J, Gackowski M, Otto F, Osmałek T. Ionotropic Gelation and Chemical Crosslinking as Methods for Fabrication of Modified-Release Gellan Gum-Based Drug Delivery Systems. Pharmaceutics. 2022 Dec 28;15(1):108. doi: https://doi.org/10.3390/pharmaceutics15010108. PMID: 36678736; PMCID: PMC9865147. ) Using this technique, sodium alginate is dripped into a calcium chloride solution to create an ion exchange that forms a cross-linked hydrogel matrix that encapsulates the enzyme. This was an appealing method as the process of making the beads is simple and the alginate material should not chemically react with the enzyme and so the risk of it denaturing is low. The internal conditions of the beads could also be optimised to match with the optimum conditions for NosZ.

Build + Test

This design was being developed in parallel with the second cycle, and reflections from both were used to inform the design of the other. As a high school team with limited time in the lab, we used a mixture of a peer-assessment method and hardware experiments in our Test phase in order to ensure we were using the most promising method.

In our peer assessment, the team working on the cell-free design passed their design to another part of the team to critically evaluate their work. This identified the following potential issues.

• If the pores of the alginate matrix are too large, enzyme leakage could occur, leading to reduced efficiency while smaller pores would decrease N2O diffusion.
• The hydrogel structure can hinder substrate movement, particularly N2O diffusion, thereby limiting the enzyme's access to its substrate. This is further supported by the fact that the diffusion of N2O through alginate or similar hydrogels is significantly lower than that of air potentially limiting the overall reaction rate.
• Estimated diffusion rates suggest a significant drop in N2O transport efficiency: ~1×10⁻² mol/m²/s in air vs. ~1×10⁻⁶ mol/m²/s in hydrogels.
• Deactivation of the enzyme during immobilisation (a common problem during all types of immobilisations) To prevent this, no chemical reactions between alginate and NosZ must happen. Unfortunately, the negatively charged carboxyl groups (COO-) in alginate could interact with positively charged residues in NosZ or interfere with the metal ions in its active site, potentially altering its activity. This could be improved by blending the beads with clay. Previous studies used Fourier Transform Infrared Spectroscopy (FTIR) to find a strong interaction between clay and the alginate bead improving stability. (note: Siti Noraida Abd Rahim, Alawi Sulaiman, Fazlena Hamzah, Ku Halim Ku Hamid, Miradatul Najwa Muhd Rodhi, Mohibbah Musa, Nurul Aini Edama. (2013). Enzymes Encapsulation within Calcium Alginate-clay Beads: Characterization and Application for Cassava Slurry Saccharification. Procedia Engineering, Volume 68, Pages 411-417, ISSN 1877-7058. doi: https://doi.org/10.1016/j.proeng.2013.12.200. )

Hardware testing performed a pilot study on diffusion rates into alginate beads. Working in collaboration with the team from Hardware and Modelling, we concluded that diffusion rates were slow enough that we should consider other options. See Hardware for more details.

While studies have shown that bacteria involved in N2O reduction can be successfully immobilised in alginate beads without affecting their physiological activity, there is no evidence of this also being possible with NosZ on its own. (note: Toshikazu Suenaga, Ryo Aoyagi, Nozomi Sakamoto, Shohei Riya, Hidenori Ohashi, Masaaki Hosomi, Hideaki Tokuyama, Akihiko Terada. (2018). Immobilization of Azospira sp. strain I13 by gel entrapment for mitigation of N2O from biological wastewater treatment plants: Biokinetic characterization and modeling, Journal of Bioscience and Bioengineering, Volume 126, Issue 2, Pages 213-219, ISSN 1389-1723. doi: https://doi.org/10.1016/j.jbiosc.2018.02.014. ) We also questioned how N2O was supplied in these previous experiments as gas and dissolved form have different diffusion rates. Additionally, the price of the alginate beads was also a consideration as using pharmaceutical-grade materials for bead production was estimated at £1366.73 while using non-pharmaceutical grade materials was £273.68. The risk of enzyme deactivation, leakage, and changes to the internal conditions meant that these prices could also potentially be higher.

Learn

After our Test phase, we engaged in a reflective process as a biology team, looking at what we had learnt, and how we could improve the project.

Given the difficulties involved in maintaining the optimal protein oxidation states, and the cost implications of this option, we decided that it would be better to focus on an in vivo system. This helped us get closer to the goals set by ourselves and stakeholders of limiting costs to ensure affordability, whilst also keeping the scope of lab work achievable so that we could make a meaningful contribution to future iGEM teams.

We were able to reuse a lot of the hardware design involved in this first cycle of design, and lots of the background research gave the team a firmer grounding in the biochemistry of nitrous oxide reduction. We were also able to work on our integrated human practices, by communicating with academics in the field, and other stakeholders to work towards a practical goal.

Cycle 2 - In Vivo Design

Design

N2OR is transported to the periplasm via the TAT system, and mutations in this system can prevent proper copper incorporation (Bennett et al., 2019). Upregulating comR represses comC, enhancing outer membrane permeability to copper and increasing cofactor concentration in the periplasm, which may boost reductase activity. Since copper is a limiting factor, we hypothesise that this upregulation could accelerate N2OR function. Upregulating ComR is a practical and cost-effective strategy for raising periplasmic copper levels. Further details can be found on the design page.

Build

We constructed the parts in the lab, but the first two attempts failed. The initial failure was likely due to unsuccessful plasmid restriction. Learning from this, the technicians at KCL performed the restriction for us, resulting in more promising outcomes with white colonies; however, it still did not work. Some red colonies suggest that the original plasmid was not fully digested and sequencing indicated recombination between the plasmid backbone and the bacterial chromosome. For the third attempt, Dr Markiv started from scratch; green colonies were recorded suggesting that additional ComR could further suppress gene expression, offering a potential avenue for future iGEM teams to explore high-copy plasmids with the ComR CDS.

For more information, please see the Results and Notebook pages.

Test

As a high school team with limited resources, we were only able to secure one week in the lab. We were therefore unable to test our parts as we had planned. However, we have detailed our Test plans to help future teams develop on this cycle of design.

• Transformation of both parts in two plasmids with compatible backbones and assessing relative fluorescence with and without the ComR plasmid to determine binding efficacy of the ComR to the binding site

• Use of a Lux copper detection system utilising CopA promoter as a possible exploration avenue for future iGEM teams. (note: Mermod, M., Magnani, D., Solioz, M. and Stoyanov, J.V. (2011). The copper-inducible ComR (YcfQ) repressor regulates expression of ComC (YcfR), which affects copper permeability of the outer membrane of Escherichia coli. BioMetals, 25(1), pp.33–43. doi:https://doi.org/10.1007/s10534-011-9510-x. )

• Utilising a gel shift assay to analyse the binding of ComR to the ComR binding site in the ComC promoter region (see design)

However, some qualitative testing was still possible based on the results of transformation. The presence of green colonies indicates clearly that the ComR already present in the DH5a E. coli was insufficient to inhibit expression of the GFP. This indicates there is an opportunity for greater expression of ComR in order to further reduce expression of downstream genes.

Learn

From our work in the lab, we took the following reflections on the construction of our parts:

• We had difficulty in constructing the parts, mainly through incomplete digestions. This could be due to a variety of factors including our lack of experience working in a lab setting or problems with the restriction enzymes used.
• There appears to have been a frameshift mutation in the ComR insert in psB1C3 (but not in psB4K5). As of the wiki freeze, analysis of the sequencing is ongoing to try and find the source of this error.

On the project administration, we also learnt the following:

• In the lab, it is best to focus work on a small achievable goal when time is limited. Therefore, having discussed with Dr Markiv, we decided to focus our efforts on building our parts to work within the iGEM framework.
• Planning for wet lab work should begin as early as possible to maximise potential time in the lab. We initially struggled to find connections with other iGEM teams in London, but we have now established a better relationship with KCL and we hope to collaborate again in the future.

As part of our contribution to our own COL iGEM teams in the future, we will work with the team next year in order to ensure these reflections are passed on.

Hardware Cycles of Design

Main objectives for hardware:

• N2O is successfully dissolved into medium with N2OR enzymes in an optimised system

• Bacterial population is sustained and carefully controlled

• Safe and sustainable disposal of waste products

CYCLE 1

Design

A cylindrical design with ‘nets’ of calcium alginate beads with nitrous oxide reductase (N2OR) immobilised within, and N2O passing through in aqueous form.

Build

We put together an analogous set-up of alginate in a beaker with solution, reflecting this original cycle of design, to replicate the diffusion process that would occur in solution.

Test

We based our experiment on carbon dioxide instead of nitrous oxide, due to their similar solubilities (1.45g/L and 1.50g/L), (note: Cubaud, T., Sauzade, M., Sun, R. (2012). CO2 dissolution in water using long serpentine microchannels. Biomicrofluids, 6(2), pp. 22002-22009. doi: https://doi.org/10.1063%2F1.3693591 ) and the difficulty of acquiring and using N2O in school labs. This would cause carbonic acid to form in the water which would thus diffuse into the alginate beads. Thus, we timed and recorded the time for a dilute acid solution of an equivalent pH to carbonic acid (hydrochloric acid) to diffuse into the alginate, as a loose estimate for the diffusion rate of H+ into alginate, and thus this could be compared to N2O. Ultimately, the diffusion rate measured into the calcium alginate was insufficient to support our desired reaction rate, especially compared to the diffusion expected through a bacterial cell surface membrane.

See the Dry Lab section for the results of the experiment.

Learn

We discovered that the alginate design would not work as a surface for reaction, and to increase the surface area for reaction and remove the problem of nitrous oxide having to diffuse into the calcium alginate, we would instead implement a solution with a free-floating bacterial colony of the transformed E.coli, leading us onto the new design.

Moreover, concerns from the biological design that immobilising the N2OR might render it dysfunctional, led to there being no doubt in the team that a new design had to be developed, where the N2OR enzymes were not immobilised and could be continuously synthesised to counteract their sensitivity and denaturing, further leading us onto a free-floating live bacterial design.

CYCLE 2

Design

Our second cycle of design holds a solution containing the population of transgenic bacteria. It incorporates a static archimedes screw design along the vertical axis, which slows the rise of N2O bubbles, therefore increasing their time spent in the solution for dissolution and subsequent breakdown. The gas is introduced at the base and the bubbles are then guided to follow a helical path around the screw before reaching the surface, where the products leave through an exhaust pipe. In the centre of the screw there is a heating element, which maintains a consistent temperature for increased enzyme efficiency.

Test

Another consideration of the live design was the requirement for homogeneity of nutrients and N2O for bacterial survival and optimal N2O breakdown rate. The purpose of our testing was to evaluate a continuous filter solution compared to another potential idea, a compartmentalised asynchronous design, in which N2O is first collected and stored inside the container, and only then does the processing of the N2O begin. Our tests were performed with openFOAM vector field modelling:

When testing the efficacy of a compartmentalised asynchronous design in comparison to a continuous filter, the rate of reaction is the most significant factor. We increase the efficacy by increasing the time N2O spends in the machine and by mixing the solution. The primary drawback of the continuous method is that bubbles of N2O travel through the length of the machine too quickly for any meaningful reduction by the bacteria to take place. A proposed solution to this was to implement a helical shaped static Archimedes’ Screw inside the cylinder.

This greatly restricts the path of N2O bubbles as they travel through the medium, allowing for a slower rate of reaction to still significantly reduce the gas. The acceleration of the bubbles upwards would be the original acceleration (without a screw design) due to buoyancy -g- multiplied by sin() where is the angle of elevation of the screw turbine. Clearly, for effects to more than one order of magnitude, a very small angle would need to be made, and thus a very long screw turbine would be needed. As this cycle of design was intended for placement in hospitals for continuous breakdown, this would be a very unrealistic adjustment, as the space occupied by the machine has to be minimised for practicality.

Aside from whether this is genuinely that effective at slowing the bubble, the effect this has on the level of mixing in the fluid is not a negligible shortcoming. The environment was modelled with (steady state) openFOAM computational fluid dynamics to demonstrate this drawback. Streamlines (the paths taken by individual molecules of water) have been drawn and coloured by vorticity. We define the magnitude of vorticity at a point P (the magnitude of the curl of the vector field at point P) to be the degree to which the fluid appears to rotate about point P.

Due to the restrictive shape of the helix, it acts to possibly increase laminarity of flow, where laminar flow can be thought of as undisturbed, smooth flow. Its regular shape reduces the degree to which it can act as a static mixer and subsequently streamlines show little overlap and overall vorticity in the structure is low.

The shape of the screw guides fluid flow throughout the structure, making the filter resistant to externally-produced body forces – particularly rotational and other forces that would against the spiral motion if the helix were not present. It follows that an impeller or other similar mixing device would not effectively increase the rate of reaction inside the environment.

It is worth noting that this applies to the fluid within the structure as that is what we are interested in mixing: that is where the bacteria are. As fluid leaves the structure (in the example of fluid continuously flowing through the filter) there is increased rotation which would result in slightly turbulent flow. This is to state the consideration of a vertically mobile fluid, although that was discarded prior to testing of our static design.

Combining this with the fact that, as experiments have shown, the bubbles of N2O are not slowed down as much as is necessary, it becomes reasonable to look for another solution.

Learn

From our second cycle of design, we learnt that a method of attempting to slow the rise of bubbles in the reducer would likely be ineffective. Based on our testing, the archimedes screw did not sufficiently decrease bubble velocity and inhibited the mixing of the solution. A variation on this design was considered, where the bubbles would pass through a series of meshes, taking advantage of the surface tension of the bubbles to delay them at each mesh, which would allow for the mixing of the reduced contents. However, this idea was also discarded, as we found that the turbulence in the solution from mixing overcame the surface tension forces, causing the bubbles to rise faster than intended. As a result, both static and dynamic versions of the continuous filter were discounted in favour of an asynchronous reducer.

CYCLE 3

Design

This image shows a CAD model of our third and final cycle of design. It features two completely separate compartments, each with its own motor to drive the mixing paddle. This allows N2O to be collected in one compartment (via a valve), while the other compartment is simultaneously processing a previous batch of N2O. Once a processing cycle is over, the role of the compartments can be switched, allowing the continuous processing of N2O. This design allows for non-continuous processing of N2O, and so can be placed outside hospital environments, removing the main area of concern for our previous designs: infection risks for patients using the machine directly.

At the front of the compartments are quick release couplings that can be used to instantly switch which compartment is receiving the N2O, requiring minimal effort by staff using the device. In a commercial version of our reducer, the switching of the N2O supply could even be automated with a simple 3-way valve. The top of either compartment features a shut-off valve, either manual or automatic. This opens the compartments, allowing the resulting gases from the processing to be released.

An important consideration for this design was the pressure buildup inside the boxes. As N2O is supplied to either compartment, the pressure inside it will build. As the box will be made out of acrylic, a strong material able to withstand pressure, this poses no problem in theory. However, in order to prevent a dangerous buildup of pressure inside the box, the back of both compartments feature a pressure relief valve that automatically releases pressure if it rises above a certain safety threshold. This can also act as an entry point for pressure, pH and temperature sensors, or, when needed, to remove and process bacterial waste from the solution.

The compartments will also feature a variety of sensors, including the aforementioned N2O concentration sensor, as well as a pH, temperature and pressure sensor. This will allow us to monitor the internal conditions of the solution, very similar to how a fermenter operates, to ensure that the conditions are optimal for the enzymes to work at their maximum rate of reaction. Although not featured on the CAD design, the entire reducer will be covered in a heating blanket that will control the temperature of the solution.

The benefits of this final design are numerous compared to its predecessors. Firstly, its construction is significantly simpler, not requiring many moving parts, nor the difficult to manufacture archimedes screw considered in the previous design. However, the real benefits of this design stem from its flexibility. As the N2O is reduced asynchronously in a sealed container, a reducing cycle can last as long as needed, controlled by sensors within the box that measure the concentration of N2O. This ensures that the processing continues until all of the N2O has been reduced, whereas with previous designs, once the N2O passed through the filter, there was no way to avoid leftover, unreduced N2O. A compartmentalised design also has the benefit of being easily scalable. Further compartments can easily be added to the design, allowing more N2O to be processed simultaneously.

Build

The prototype of our final design cycle was made predominantly using laser cut acrylic, as well as some 3D printed components, including the paddles and motor mounts. The purpose of our prototype was to test the ease of manufacture of the design, and evaluate the potential cost of building a single unit.

Test

To compare to the previous cycle of design, the same openFOAM analysis was carried out for the new design, modelled as a cylinder. Physically, we suggest a box due to the simplicity of assembly compared to a cylinder, though the models will be roughly analogous:

The benefits of an asynchronous reducer are simple: the machine can continue for as long as is necessary to reduce the N2O to a satisfactory concentration. Seeing as there is no continuous filter-like component to the machine, the primary container of reaction can take a shape that facilitates mixing more readily. A cylindrical mixing container with a single impeller was modelled turbulently. The following pictures have been taken of the velocity field at an arbitrary time during the mixing process.

Qualitatively, the vectors are clearly less ordered, indicating a more mixed solution. Quantitatively, the vorticity has increased by an order of 105. The streamlines demonstrate much of what can be inferred from the velocity field and the magnitude of vorticity, a stark change from the arranged streamlines in the helix.

In conclusion, due to the limited effect the helix has on bubble velocity in the filter and the extreme detriment it has on mixing of the fluid, both static and dynamic versions of the continuous filter were discounted in favour of an asynchronous reducer. The design will now be fully introduced.

Learn

Our final design prototype was easy to manufacture and relatively cheap, only costing about £160 (See below). However, the difficulty in making the container airtight was greater than anticipated. After significant effort, we managed to make the container waterproof, but we were not able to test whether it is airtight, a much more difficult task. Our prototype therefore showed that the laser cutting method of building the box may not be viable in a commercial product, due to the unreliability in making it airtight. Instead, a vacuum moulded container could be used, which can be consistently made airtight. Furthermore, our prototype did not include any automation or sensors discussed earlier. While these factors might increase the projected cost of 1 unit by some amount, our hardware is significantly cheaper than alternatives currently on the market, by 3 orders of magnitude, and with an established supply chain, the cost per unit could be lowered even further than our estimate.

Thus, for future teams, we would propose investigating the automation of the valve-switching, pressure-monitoring and bacterial waste filtering, which currently would require human maintenance.

Projected Cost:

Total for parts ordered: £107.96

These include: 12V motor, 12V battery, 2x switch, Temperature and Pressure Sensor, Shaft Coupling, Shaft, Epoxy, 3m gas tubing, One way valve, Shutoff Valve, 2x quick coupling, 2x Waterproof bearing, 2x heating element (projected cost: £25), Insulator (projected cost £20)

Total (including heating element and Insulator coating): £152.96

Projected total cost:~ £160

Construction of the final proposed model:

Please find attached a manual to assemble the final hardware cycle of design: