The most important tool for our success this year was the engineering design cycle. When working on a project of this scale, many times it’s tough to know where to start and how to successfully develop ideas and get going with an idea. That’s where this process really helped us hit the ground running right from the start, and it helped guide our decision making, planning, idea generation and more. The design process consists of seven steps. In order, they are define, research, imagine, plan, create, test, improve.

Following this process led to many different iterations and designs throughout the project. As new challenges and issues arose, our design needed to be re-evaluated in order to meet the new criteria in both the pathogen detection side of the project and the filter portion. We had to constantly take our stakeholders into account and really figure out what was best for them and what they needed from our work, as well as always keep ethical considerations into account.

Over the course of our project, we followed iGEM’s Engineering Design Cycle of Design, Build, Test, and Learn in multiple cycles. In terms of our Bst DNA polymerase fusion, three pivotal engineering phases led to the final design: (1) fusion with the terminal protein region 2 (TPR2) subdomain of the polymerase ϕ29, (2) fusion with DNA binding protein Sso7d and (3) final fusion with DNA binding protein Sac7e.

TPR2 subdomain of ϕ29 plays a critical role in the enzyme’s processivity. Fusion of this region to Bst was predicted to have similar effects.

In silico building of this fusion protein brought up concerns, as the structures of ϕ29 and Bst polymerases greatly differ, and key residues of ϕ29 lie in different domains in Bst. These findings made the ‘Test’ portion of the design cycle negligible.

From this preliminary design, we learned that base structure similarity would provide us with the greatest chance of success when it comes to the engineering of our fusion protein. 

Previous research found improved polymerase processivity upon fusion of Sso7d to Taq polymerase (1). Given the structural homology of Taq to Bst, we hypothesized greater chance of success through this fusion. Design of a flexible (GGGGS)4 linker was inspired by a similar linker in another study involving Sso and Tzi polymerase (2).

Our modified Bst polymerase was generously supplied by Integrated DNA Technologies (IDT) as a gBlock. Initially, we expressed the protein in a variation of pET16b vector but were unable to complete expression this way. We overcame this issue using pET24d instead.

Successful expression and purification were confirmed through SDS-PAGE and a Western blot.

This design cycle taught us the skills and protocols required to express and purify a protein. 

Sac7e is a homologue of Sso7d but has greater binding affinity to double-stranded DNA (3). Inspired by previous research that made mutations to Taq, we chose to make point mutations in Bst in hopes of improving the enzyme’s thermostability to account for fluctuations in temperature in a point of care device (4).

High similarity between Sac7e and Sso7d meant that the same (GGGGS)4 linker could be used. Determination of which mutations to make came from in silico experimentation using YASARA modelling software with the FoldX extension.

Based on what was learned in the previous engineering cycle, we used the pET24d vector again. According to our YASARA trials, Gibbs free energy at residues K549W, K582L, and Q584L before and after mutation measured from 0 kcal/mol (initial) to -0.126 kcal/mol, -1.63 kcal/mol, and -2.039 kcal/mol, respectively (final).

How to use new software (ex. YASARA) was learned during this final phase of engineering our modified Bst DNA polymerase. While we were able to fully express this new protein, the purified sample was not seen on the Western blot. Therefore, difficulties involving purification methods (in our case, nickel chromatography) were also observed.

Our kettle device used multiple iterations from start to finish. The original idea for the filtration portion of the project was intended to be an on-tap filtration system. This device would screw on to any faucet at home and work to filter out heavy metals and bacteria just like the kettle does now. After design changes and consulting with stakeholders, our goals shifted which meant our project did too, which led us to eventually design a kettle. Throughout the entire process, the design of each part had to be sketched, 3D modelled using CAD softwares, and toleranced for manufacturing or 3D printing. Each part had to be able to fit into another tightly without leaving too much space in between or not enough space causing an interference fit. The temperature sensor for the kettle was also modelled using Tinkercad before purchasing and building our circuit. We used the software to predict the voltage outputs from our microcontroller, as well as to ensure the circuit ran without any shorts or breaks.

Once each CAD model was designed, we would then have them 3D printed and connected to other parts to make sure they fit together snuggly using mainly nuts and bolts. Due to the inaccuracy of some vertical 3D printed holes, some prints would need to be scaled slightly larger to leave enough room for bolts to fit through. We would also physically measure each print down to a hundredth of a millimetre to ensure the next part we created would be accurate with previous part. Circuit components were assembled using mid-sized breadboards and electronic components were wired using jumper wires.

Once parts were made and assembled, we would physically test parts to help make improvements. Three main tests for the kettle came from the heat plate, the temperature sensor, the mechanical filter and the kettle.

The heat plate first was tested using basic thermodynamic principles, and we used this to predict how hot the water would get from the plate with the given thickness of the kettle’s base, the power of the hot plate, the material of the kettle (aluminum), and the volume of water. Once we modelled this and made sure it would be possible in theory, we moved on to physical tests using our heat plate with our aluminum kettle and compared the results.

The temperature sensor was built and ran using the Arduino IDE controlled by an Itsy Bitsy M0 express. The circuit was tested using the microcontroller to make the LED’s turn on and off when the temperature reached a certain point. During said tests, we had to compare the temperature with an actual thermometer to calibrate the senor to get accurate readings.

The mechanical filter and kettle were parts we couldn’t predict using software and required us to physically test it for the best results. To do this, we would run 1 cup of water through the filter and recorded the time it would take to fully clean the water. For the kettle, we took 500 ml, 1L, and 1.5ml of water and recorded how long it took for the water to reach its boiling point.

Once we had tested all aspects of our kettle theoretically and experimentally, we would then look for ways to improve upon our design to make it smaller, faster, tighter etc. For the kettle this involved new CAD models, sketches, ideas, and testing different solution to the same problem.