Team:Calgary/Engineering
INTRODUCTION
The engineering design cycle consists of four major steps: design, build, test, and learn. This year, we were committed to ensuring we followed the engineering design cycle for each of our major projects in order to optimize them for the best performance in the future. While this page provides an overview of the engineering design cycle for each of our subprojects, please visit the individual pages to learn more about the projects.
PROTEIN PRODUCTION
Throughout this iGEM season, the Neocycle team has been working on several subprojects, two of which, metal separation and measurement, required wet lab synthetic biology techniques. For metal recovery, we worked on designing lanmodulin (LanM) constructs fused to a cellulose-binding module. This served to hopefully immobilize LanM to an adsorber column such that its function could still be preserved. The two cellulose-binding modules we chose were CipA and Cex, each chosen for their relative ease of expression in E. coli and high affinity for cellulose immobilization. These modules were chosen after discussion with our HP contacts, who helped inform our construct design.
While metal recovery was looking for cellulose-binding modules, the measurement subgroup was working on developing three different systems - BRET, Lucifer, and Elektra. Similar to metal recovery, the systems involved LanM fused to another protein. For BRET and Lucifer, this was a two-protein complementation reporter. For Elektra, this was a signal molecule attached to one end of LanM, and a covalently-bound gold plate on the other end. All of these designs were based off of previous literature searches, discussions with HP contacts, and feedback during the design process. Some of our events, such as our Faculty Talk, where we pitched the Neocycle process to industry contacts, professors, and general iGEM enthusiasts, helped us inform our design and begin building constructs.
Initially, both subgroups ordered constructs in the pucIDT vector, a vector from IDT used to store large DNA sequences. While initial transformations of these vectors were successful, we repeatedly found ourselves getting stuck on the protein production and purification phase of the project. As we went through several troubleshooting steps, such as changing the induction method for our protein, changing the protein extraction and purification methods, we found that we were continuously getting stuck on the protein extraction step. Our tests of the design weren’t working. As such, with advice from our HP contacts, we decided to go back to the drawing board and re-order our constructs in g-blocks.
This time, we decided to change vectors and clone our constructs into the Xpress vector designed by the Schryvers’ lab. This vector was unique due to two main reasons: first, it contained a Glutathione S-transferase tag (GST tag), which is ideal for expressing difficult proteins. Second, it contains both an ampicillin and kanamycin resistance gene, allowing for quick and easy screening. Upon successful cloning into the vector, the kanamycin resistance gene will be removed, but the ampicillin will remain. As such, successful constructs will grow on only ampicillin, while unsuccessful constructs will grow on both kanamycin and ampicillin plates.
After successfully cloning our WT-LanM-GST construct into the Xpress vector, we were able to successfully produce, purify, and test the efficacy of the protein in a metal binding assay. While we experienced multiple iterations of the engineering design cycle, future directions include producing, purifying and testing the LanM-fusion constructs, as well as comparing their activity to WT-LanM.
Read more about our metal recovery project here and our measurement project here.
BIOLEACHING
In order to put metal separation into practice, we knew that we would have to find a way to solubilize the REEs out of e-waste so that they could be accessed by LanM in solution. This can be done using synthetic strong acids, but this carries environmental and safety hazards, so we turned to bioleaching. We sought to demonstrate that we could use G. oxydans to generate acids and leach REEs into solution from e-waste. We chose to demonstrate that bioleaching with G. oxydans is effective in bioleaching two types of waste (NdFeB magnets and hard drive shredding dust under our conditions). The bioleaching team set up two rounds of experiments with different concentrations of metal, types of feedstock, and incubation time. From these experiments, we learned that G. oxydans is capable of successfully leaching up to 82% of total neodymium present when incubated with 10 g/L NdFeB magnet for 7 days under our conditions. We also were successful in leaching a total Nd content of 9.6% total mass from hard drive shredding dust. To learn more, check out the bioleaching subproject here.
ELEKTRA
Developing the construct for Elektra required intensive literature research to ensure the LanM protein would be covalently attached on one end to the gold electrode and on the other end to the ruthenium signal molecule. In order to immobilize one end of the LanM protein onto the gold electrode (to measure the oxidation current) cysteine amino acids will be utilized. There was literature on protein immobilization using a cysteine amino acid [1]. By using this technique, the protein is able to create a strong covalent bond to the gold electrode and self-assemble into a dense protein monolayer that could be utilized for this measurement system [1]. On the other end of LanM, there was literature summarizing all the ways a ruthenium signal molecule could be attached onto a protein in order to develop biosensors [2].
Based on the target amino acid, the protein can be treated in a series of steps in order to covalently attach the ruthenium molecule [2].
Since the system depended on the ruthenium signal molecule to only be added onto one end of LanM we had to attach the nanoparticle to a chemical group that was unique to the n terminal end of LanM. When scanning the sequence of LanM, it was determined there were no histidine residues in lanM, which makes this amino acid a strong candidate for the ruthenium signal molecule immobilization.
LUMOS
The primary purpose of Lumos is to be a hardware designed to help individuals out during the REE extraction process. As such, it can be said to be a human-centered design, with ease of accessibility being a primary point. The Lumos hardware went through several iterations of the engineering design cycle before becoming the product you see today. Initially, Lumos was based off of previous iGEM fluorometer designs, but this was soon found to be too insensitive to changes in light for our purposes. As a result, it was necessary to change our light source. For more information on the engineering design cycle and how several iterations helped inform the Lumos design, please visit the Measurement Hardware page.
METAL RECOVERY
BIOREACTOR
We applied the engineering success design cycle to the development of our bioreactor prototype LanHome. The goal of the process is to design and develop a low-cost bioreactor that allows for upscaling protein production at a laboratory scale in order to investigate potential industrial upscaling. With the first step of designing the bioreactor, we needed to first define the need criteria of the device, which we defined as the following: aseptic culture, sterilizability, mixing capability, and capacity for sampling and monitoring, and low cost. With the need criteria defined, we explored different designs that would satisfy the needs and decided to go with a simple and effective prototype set up that would allow for further testing.
Once we drafted a design for a fed-batch stirred tank bioreactor system, we created a material list and determined the material we have access to in the lab setting and ones that need to be purchased in order to effectively assemble a cheap prototype for testing. After that, we entered into the build stage of the engineering design cycle, and with the materials all collected, we assembled them based on our design and modified areas that were different from what was envisioned. Once the hardware was assembled, we entered into the implementation stage where we tested the hardware’s capability to match the needs outlined by inoculating the reactor with E.coli and collecting OD measurements to validate its ability for aseptic culturing. The results demonstrated that it did meet the need criteria defined, however it did not outperform conventional shaker flasks. The next step is to use the results to identify parameters that can be modified such as oxygenation rate and mixing capabilities to further refine the design in order to optimize the process. The cycle returns to the design stage where once we implement the iterations into the design we can build it into the system and test it again.
ADSORPTION COLUMN MODELLING
With PACLan, the adsorber column used for recovery of REE ions, we approached the design process with more caution through the implementation of modelling and simulations as the build and testing steps are both reagent and material expensive. Through defining the operation criteria we wanted to achieve in order to meet industry standards, we began the modelling of the changes in the different process parameters that would allow us to optimize our process.
Through ideal conditions and physical principles underlying mass transfer, we designed a system of partial differential equations (PDE) describing PACLan under ideal conditions. We then implemented it through a numerical simulation using the finite-difference approximation to reduce the system of PDEs into a system of linear equations. We implemented the model using baseline literature values and swept through different values of parameters in an attempt to learn more about the behavior of the model output through performance parameters. After receiving simulation data, we described the trends, analyzed them, and proposed explanations for the observed trends. After realizing that our model may not be valid under realistic conditions, we accounted for non-ideality as represented by axial dispersion in the column.
Using the same steps as before, we implemented the corrected model and probed its response by varying the parameters and seeing the changes in the output. We then compared the trends in the ideal and non-ideal cases and chose the parameters that maximize the processing capacity of the unit through our analysis of the simulation data. The whole cycle will begin again by adding more complexity to the model, implementing it, and analyzing the trends which will make simulation data more reflective of real-life conditions.
Click here to learn more about the bioreactor design, and here to learn more about the adsorber column.
