Introduction

After discussing our project’s goals and future opportunities with a microbiologist he insisted that we test our system in a bioreactor. One thing is our system working in a flask but it's not given that the same system would work on a bigger scale. First step of proving our validity in a high volume system is to test the process in a lab-grade bioreactor. These tests would provide data on how to scale up our systems and to prove that our glue could be mass produced.

Implementation

A laboratory-scale bioreactor is costly, leading us to consider constructing one ourselves, with the intention that our plans could be beneficial to other teams. Initially, we outlined our objectives, specifying the desired functionalities of our reactor. We aimed for full automation to minimize human error, which could compromise data integrity. Bioreactors typically incorporate various sensors and components, so certain parts were identified as essential for the goal:

• Temperature sensor: used for monitoring the temperature of the bioreactor, ensuring that the temperature is kept at the optimal level for the bacteria to grow.
• Heating element: used to heat the bioreactor when the temperature is too low.
• pH sensor: used for monitoring the pH of the bioreactor, allowing us to adjust pH levels if necessary.
• Mixing impeller: used to continuously stir the contents of the vessel, ensuring homogenous mixing and therefore providing the cells with substantial oxygenation of the medium.
• Turbidity sensor: used instead of Optical Density (OD) sensor to measure the growth of the bacteria.

A diagram of the main components of the bioreactor can be seen below (Fig. 1):

For the container a glass jar with an aluminum lid was used, an aluminum lid was chosen because it is easy to modify and doesn’t require complex tools. All of the parts and sensors were fixed to the lid, this design choice was made in order to allow sterilization. The bioreactor must be sterilized after each use; however, the sensors and electronic components cannot be autoclaved. The sterilization process comprises two steps: first, the lid and base are separated, after which the base is placed into the autoclave. Lid components were sealed and waterproofed using a silicone based sealant. This allowed us to submerge the lid with all the components into 70% isopropanol solution, which would effectively sterilize it. Holes were drilled in the lid according to precise measurements for each component that needed to be inserted into the reactor. Each part was enclosed, sealing the bioreactor from the outside and reducing the risk of contamination. All the electronic components were fixed in a box separate from the bioreactor.

The full list of components required for building the bioreactor, including their prices, is shown in Table 1:

The bioreactor in action is shown in the picture below (Fig. 2.). The electronics of the bioreactor is placed inside the box on the left, protecting it from any external damage, also making the setup more visually appealing.

Measuring bacterial growth

Our first idea was to use an OD meter to determine the optical density of our medium. But that plan fell through when we saw the prices of OD meters and the fact that OD meters are not directly compatible with the Arduino interface. Our bioreactor is designed with accessibility in mind: we aimed to create a foundation for other teams by providing simple instructions for building an affordable, lab-grade bioreactor. For this reason, we opted not to use OD meters and after scouring the web for alternatives, we found turbidity sensors. Turbidity is a measurement of how much sediment is in a liquid. This technology is not often used in biology as a way to determine bacterial growth. Finding a workaround was not easy but we came up with an idea of making a calibration curve [1]. Measuring turbidity and OD at the same time increments gave us enough data to make a calibration curve. Plugging in the turbidity value into a simple formula (y = mx + k) gave us the corresponding OD. This process was automated to deter human error (more about that in Software wiki page).

Electronics

For the "brain" of the bioreactor, we decided to use the Arduino Uno R4 WiFi. It is a versatile, open-source microcontroller with built-in Wi-Fi capabilities, making it ideal for remote monitoring and control (more about that in Software wiki page). The Arduino platform is highly favored in prototyping because of its simplicity, flexibility, and large community support. Additionally, the Wi-Fi functionality allows seamless integration with cloud-based services and data logging systems, which are crucial for real-time data monitoring. One of its most important features is the ability to interface with various sensors, making it highly adaptable to the bioreactor’s requirements, such as temperature, pH, and turbidity sensors. Furthermore, its cost-effectiveness and ease of use make it an attractive choice for both academic and industrial applications.

All the system's electronic components, connected to the Arduino, are displayed in the schematic (Fig. 3). The turbidity and pH sensors are connected to Arduino's analog pins, while the temperature sensor, DC motor control module and heater's relay are connected to the digital pins. The DC motor and heater is powered by 12V external power supply, while other devices - by 5V power supply provided by the Arduino microcontroller.

A more detailed instruction on how to build the bioreactor and connect the electronics together can be found the document provided below (Fig. 4):

Engineering

Adherence to sound engineering principles guided our team throughout the entire project, allowing us to maximize the efficiency and quality of the final product (Fig. 5). In our approach, the "design" phase focused on problem identification, tool selection, and planning. The "build" phase involved assembling key components, while the "test" phase assessed their functionality. In the "learn" phase, we reviewed data to refine future iterations, continuously improving our designs.

3 iteration

Design

Overall design stayed unaltered but separate components were tweaked.

Build

We used a silicon based sealant for our new turbidity sensor and waterproofed it. Screw on our temperature sensor was the least of our problems, we just removed it. And the final problem of our previous design was our rusty impeller rod. We used a rubber based coating spray, after multiple coats the rod had no contact with water and was successfully sealed.

Test

Third test evaluated our sterilization method. It was run with 600 milliliters of LB medium without antibiotics and the conditions (37 celsius, X rpm’s and neutral pH) that we plan to grow our bacteria in. The bioreactor was left working for 3 hours straight then we turned it off and left it overnight at room temperature. This test came out positive because the LB medium did not get contaminated.

But a problem occurred. We observed that our bioreactors sensors were too high for the volume of LB we used and some of the sensors were right at the liquids surface.

Learn

• Need to lower our sensors for the bioreactor to be adaptable for a wider range of usable volume
• Our sterilization method has been proven effective

Validation

To validate the functionality of our bioreactor, we conducted experiments using E. coli containing our complete CB2 plasmid system (for more details, refer to the Design section of our wiki). During the first experiment, we did not yet have our turbidity sensor, so optical density (OD) measurements were taken manually using a spectrophotometer. The experiment utilized 700 mL of LB medium and 10 mL of overnight bacterial culture with an OD of 6.47.

The graph indicates a consistent growth pattern, with no noticeable dips or unexpected peaks.

The second test was conducted shortly thereafter, using the same set of CB2 plasmids under identical conditions. However, in this test, the OD was measured using our newly implemented conversion system. As a control, manual OD measurements were periodically taken with a spectrophotometer to ensure accuracy.

The second test confirmed the results of the first, as the growth rate remained consistent. No significant differences or fluctuations were observed when comparing the data from the bioreactor to that obtained from the flask-based expression. While expression in the flask and a standard incubator occurred at a faster rate, this may be attributed to the incubator's ability to provide uniform heating, as airflow ensures consistent temperature distribution around the flask. In contrast, the bioreactor’s heating source is fixed to a specific location, with the impeller being the sole mechanism for distributing heat. The impeller may be circulating the heated medium too slowly, resulting in inconsistent temperatures for the bacteria.

Additionally, our bioreactor is constructed from glass, which is not an optimal heat insulator. This could result in heat escaping from the sides of the glass container, creating two distinct temperature zones. This supports the hypothesis that the slower growth rate in the bioreactor may be due to inconsistent heating throughout the system.

The implementation and initial test of our conversion chart yielded positive results (Table 3). A comparison between Table 2 and Table 3 reveals a similar overall pattern, although the table displaying the conversion results shows some inconsistencies. Several factors could have contributed to these variations, including micro-debris in the sensor, inadequate mixing speed, or inconsistencies in the calibration graph. Despite these minor discrepancies, the results are promising and provide a solid foundation for further improvement.