Overview
What Is A Turbidostat?
To have a general understanding of our system, it is important to understand what a turbidostat is. Simply put a turbidostat measures the turbidity of a bacterial culture. This is often accomplished through optical density measurements. The more light a culture scatters from the turbidostat’s light source, the less light the accompanying detector receives and the more turbid the culture is.
Why Is A Turbidostat Important For PACE?
A turbidostat is important for Phage-Assisted Continuous Evolution , or PACE because it ensures that the bacteria the bacteriophage infects are healthy. When the bacterial culture has just started, immediate phage introduction results in metabolic strain and bacterial death. By using a turbidostat to collect optical density data, we can:
- Monitor bacterial health via optical density
- Identify mid-log growth for optimal phage introduction
- Prevent premature bacterial death
- Allow adjustment of selection pressure for phage replication
Our System
The goals of our hardware can be broken down into two categories: maintaining healthy bacteria for phage infection and having the ability to adjust selection pressure.
- To maintain healthy bacteria, we conduct a calibration experiment to determine the optical density where the bacteria has reached mid-log growth. In PACE, this set point determines when the bacteria are moved into the “lagoon,” the culture bottle holding phage.
- Regarding selection pressure, our wet lab team developed optogenetic tools to allow for the adjustment of selection pressure using light. Leveraging their genetic tools, we incorporated adjustable LEDs into our 3D-printed hardware of the lagoon and turbidostat.
To understand our system in greater detail, we recommend first viewing the Major Components section. We also utilized Cadasio, an online platform, to create a three-dimensional, interactive visual representation of our system, how to assemble it, and its key components. This model is the most comprehensive physical overview of our system; it can be viewed here. Additionally, we created a video demonstrating our system's various functions. For users intending to construct this system, after you have established a general understanding, please navigate to the Turbidostat Construction and Usage section and view the associated links.
The Problem Our Hardware Solves
According to existing literature, developing a successful PACE workflow can take upwards of two years because researchers must develop genetic constructs while constructing the physical hardware to enable continuous evolution (Popa et al., 2020). The time investment and lack of knowledge regarding continuous evolution hardware make PACE unfeasible for researchers interested in protein engineering. While there are existing commercial solutions like eVOLVER, these systems are complex, difficult to use and considerably expensive.
Through our literature review, we noted the need for easy-to-use, easy-to-assemble, affordable continuous evolution hardware. Conversations with protein engineering experts like Dr. Takahashi and Dr. Badran deepened our understanding of the time investment for initial hardware construction, setup, and cleaning, and of the potential challenges of biofilm accumulation.
Ultimately, our team identified a clear need for simple, quick-to-assemble, user-friendly, and reliable PACE hardware, aligning with our overarching goal to make protein engineering more accessible. To clearly define our goals, we broke down accessibility into four main categories: creating a turbidostat that is easy to use, affordable to construct, minimally complex, and easy to modify. Each of these goals had to be met while still maintaining the system’s core functionality and ensuring reliability.
Major Components
The following is an overview of the system’s major components* and their functions. To view a complete 3D model of our system you can view our Cadasio assembly and user instructions here and select “System Overview”** from the menu.
Stock Media: 1000mL corning bottle that holds media
Turbidostat: 100mL corning bottle that cultures our bacteria until it is ready for the lagoon. 3D-printed parts hold the culture bottles and remaining electronic hardware
Laser: Sends light through our culture to the photoresistor
Photoresistor: Receives light from the laser outputting a value to the microcontroller, which is used as our optical density (OD) reading.
Stir Motors: To keep our bacteria healthy and guarantee consistent optical density measurements, the culture is spun by a CD motor with a magnet attached, moving a magnetic stir bar inside the bottles.
LEDs: LEDs placed on the inside of our 3D printed parts allow for the adjustment of selection pressure through optogenetic control.
Control Unit: Contains all of the electronics to operate and interact with our system, including a 3D-printed case to hold components.
Microcontroller: The Microcontroller or Arduino is the programmable circuit board that executes our system’s code.
Printed Circuit Board(PCB): Connects all of our essential electrical components to our microcontroller.
Rotary Encoder: Allows the user to navigate the LCD menu and select items including changing LED color, selecting the type of experiment, and cleaning.
LCD: The LCD screen displays all the relevant information for our system and its operation.
Lagoon: 100 mL corning bottle that contains the bacteriophage, and cultured bacteria strain after it reaches mid-log growth. 3D-printed parts hold culture bottles and the rest of the electronic hardware.
Stir Motors: Same function as the stir motors in the turbidostat.
LEDs: Serve the same function as in the turbidostat.
Peristaltic Pump: The peristaltic pump moves media and bacteria through our system.
Waste Media: 1000 mL corning bottle which collects all the waste media from the system.
Additional Necessary components
Laptop: In our current version, a laptop is necessary to upload the code to the Arduino and record the optical density values. In the future, we hope to write code so that the optical density values are stored on an SD card.
Incubator: The system needs to be placed inside an incubator to keep the bacteria cultures at the right temperature.
For further information, regarding our previous and current design iterations, reference the Hardware Design Build Test Learn section on our Engineering page.
Design Approach
Inspiration for our design: While our hardware makes some key innovations on existing turbidostats, we would like to acknowledge that inspiration for many parts of our design and valuable mentorship was provided by many individuals in the Bartelle lab, including Dr. Benjamin Bartelle, Fernando Flores Mora, Ashley Tse, Alpha Sinworn, and Gabby Cerna. Their work and guidance helped guide our decision to use a laser to monitor optical density and inspired our design of modular silos to contain each culture flask.
Our design approach to making our hardware accessible can be broken down into four main categories: Usability, Affordability, Complexity, and Modification.
Usability
The effective implementation of hardware in a lab necessitates a smooth user experience. We aimed to address the two critical barriers regarding this: the barrier to implementation (the time, expertise, and problem-solving skills to create a functional system); and the barrier to continued use (maintenance and reusability).
Actions:
- Provided Cadasio Assembly and User Instructions to help users with initial system setup and continued use
- Provided all system information for the user to 3D print and modify as needed
- Provided code that requires little user modification beyond entering the system set-point from the calibration experiment
- The settings in our “LCD Code,” with the rotary encoder and coupled LCD allow for the operation of the pump without interacting with any code
- Luer Lock connections for the tubing of our system allow for easy disassembly for cleaning
Affordability
Excluding the cost of the tools required in assembly, we set a goal to have our total system components cost below two hundred USD. Our final system costs approximately $124. Additionally, we ensured our design utilized tools commonly found in university labs. To increase affordability we reduced the number of components in our system while still maintaining usability, tunability, and overall function.
Actions:
- Reduced to one peristaltic pump per system, eliminating an additional cost of roughly twenty-five USD per additional pump
- Created a system that was more than $50 less than our initial intended component budget
- Created custom caps to save on cost
- Used widely available inexpensive components
- The system is modifiable to the components that users have on hand, potentially decreasing the cost
- The simpler lite version of our system was created for increased affordability
Complexity
The complexity of a system can make it challenging to set up, burdensome to use, and impossible to rely on. While a complex system is correlated to higher costs, removing too many components can diminish system utility. Therefore it was imperative to determine how to make a simple system and maintain core functionality.
Actions:
- One peristaltic pump reduces the complexity regarding the code, priming, and set-up compared to a typical PACE system with two or three pumps
- Provide “Lite Code” and “LCD Code” for the user to determine the degree of complexity
- Mapped all the functions of the system onto one button rather than several
- Connected all the stir motors to reduce wiring and PWM pin usage
- All the lights are controlled from a single data line through the use of NeoPixel LEDs
Modification
Our hardware is a foundational design that our iGEM team and other labs can modify, build upon, and customize. Depending on the user's intended use, they can alter the turbidostat, code, and 3D part design. We intended to make our hardware easy to modify to foster the continued innovation of our system.
Actions:
- The 3D design on Onshape can be edited to any size the user desires
- System Calibration allows the user to choose any optical density measurements correlated to their preferred growth stage
- Annotated code allows for better understanding and easier modification
- Existing LED colors allow the system to fit user needs
- Detailed documentation of circuit design allows for easy modification
Demonstrating System Utility and Functionality
To show the functionality of our system we sought to demonstrate the system’s ability to culture bacteria to mid-log growth and then move the bacteria from the turbidostat to the lagoon once they reached mid-log growth.
Interpreting the Data:
During our experiments, adding fresh media to the turbidostat washes out old bacteria, increasing the light intensity, whereas bacterial growth decreases the light intensity. With more light, the resistance across the photoresistor decreases, increasing the voltage reading and decreasing the optical density.
As reflected in our calibration experiments, the voltage readings of the photoresistor follow a roughly logarithmic curve- demonstrating bacterial growth. Similarly, when running a PACE experiment, the voltage readings follow a similar curve until they reach the set point, where media is added, slightly increasing the voltage reading and decreasing the optical density. The bacteria then grow, decreasing the voltage reading and increasing the optical density until the sensor reaches the set point again, and the process repeats. The system should effectively keep the optical density of the culture stable concerning the set point. To conserve lab resources, each of our tests was conducted with 400 mL of media, often running out since we ran our tests overnight. When the media ran out, this resulted in erratic voltage readings, which are irrelevant to the success of the system. Additionally, our later graphs include a sharp spike to -399 when the peristaltic pump had been turned on, providing evidence that bacteria were sufficiently flowing out of the turbidostat to the lagoon.
-
06/03/2024 - Our first attempt at a turbidity experiment that was run without pumps and simply measured the OD of our bacterial culture over 13.7 hours. The average values were calculated every minute. -
07/05/2024 - We attempted to use the pump to move the media. The pump ran when the OD value was less than the set point, which we determined from the previous experiment. Optical density values were recorded every three seconds. The results were very erratic and less than ideal. -
07/09/2024 - This time we ran the system with just water instead of media or bacteria to ensure that our components were measuring OD value accurately as the previous experiment on July 5th resulted in a graph with a large range of values and no clear growth curve. This was also when we started graphing -399 when the pump was on. -
07/13/2024 - Instead of having the code check the OD value that is taken every 3 seconds, we instead measured the OD value every second and changed the code to run the pump based on the averaged value from every 10 seconds. Additionally, rather than connecting the laser to a 5V power supply, we utilized the Arduino 5V power pin. We suspected that fluctuations in the power supply may have caused fluctuations in our readings. -
07/14/2024 - The data from the test utilizing pumps and media was largely faulty. After switching the 3D printed hardware, our photoresistor was damaged and replaced with the wrong model photoresistor. It was also held in place using tape. The photoresistor likely shifted during testing, resulting in a dip in values contributing to the inaccuracies. -
07/15/2024 - In this test, bacterial growth followed a normal curve before leveling at a value above our set point this meant the pump did not run when the OD leveled off and as a result, the stock media bottle had significant media remaining. -
07/18/2024 - Due to the media remaining in our stock media bottle in the previous test, we ran an additional calibration experiment, our assumption was that our set point was too high and the bacteria died prematurely, to test this we wanted to recalculate the set point. -
07/23/2024 - This test was performed with the new set point calculated from the previous calibration experiment. -
07/31/2024 - To ensure that our data was consistent from one test to the next, we did not change any set-up aspects including the set point, hardware, media, and bacteria. Past two hours, the media ran out, resulting in the pump continuously running but prior to that we had similar results. -
09/23/2024 - Our last experiment was to ensure the LCD screen and rotary encoder operated as intended. The whole experiment was run by only selecting “Run PACE” and “Start.”
User Testing
We performed two rounds of user testing where we asked volunteers to set up a mock PACE experiment using tap water as a proxy for media and bacteria. We had them follow our Cadasio Assembly and User Instructions, navigate to our GitHub, upload the code, and operate the system using the rotary encoder. All of our volunteers were undergraduates who participated in ASU’s DIYbio club- we chose to do user testing with these students as we feel they closely represent our end users who we hope to be undergraduate or graduate researchers who have an interest in protein engineering but minimal hardware knowledge/experience. We took notes during the testing and helped them through any expressed confusion.
Insights from the First Round User Testing:
- Need to improve instructions regarding uploading the code the link did not open a new window and the project step the user was on was lost
- Necessity of providing instructions for how to use the rotary encoder to navigate the menu
- The user provided feedback on how to clarify instruction terminology and color-coding aspects of the system
- The user suggested Viton tubing for added durability
Changes made between rounds:
Between the user testing rounds, we were able to make a series of small changes such as grammatical errors, instructions for rotary encoder usage, and improvements in initial system setup. Additionally, during our second round, we ensured the alignment of our physical hardware components mirrored our digital model to reduce confusion.
Insights from the second round:
- The model requires clarification on which pipettes to use when connecting tubing
- The digital model needs to further differentiate between what pipettes are meant for air and which are meant for media
- The user suggested combining certain steps of the instructions for further clarity
- The user expressed slight confusion when determining if the system was running and if the rotary encoder must be held down
- Better background information about the system
Future Plans
Our system's future development could go a multitude of different ways. Thus, we hope this project provides a platform for other researchers to develop additional novel functionalities. Initially, we had hoped to add the capability of measuring waste luminosity to monitor phage infection and considered adding a heating element with temperature feedback so the system could run independently of an incubator. However, our future plans currently center around short-term goals to increase reliability and accessibility to make the most intuitive platform for others to build upon.
CAD:
For users who wish to modify culture bottle sizes, we aim to make all dimensions variable-driven, so a user can enter their required dimensions and the system size will change without having to interact and find and edit each dimension within the CAD model. We also aim to utilize Onshape variable tables to set up a series of common configurations of culture bottle sizes the user can select without editing dimensions.
3D Parts:
We have found a few reliability issues with the snap-to-fit parts that we hope to improve so they securely fit together while also being easy to disassemble. For example, the rotary encoder cover occasionally gets stuck if pressed too far in, requiring a screwdriver for removal.
Electronics:
While we were able to design a PCB for our system we did not have time to integrate and test the PCB, this as well as refining other small elements of wire management and connection are part of our future goals.
Code:
While we came close to successfully implementing the SD card to store optical density values, we ran into code issues that we did not get a chance to debug. By implementing an SD card, the user does not have to leave a laptop connected to the system during the entirety of the experiment.
Additionally, we plan on changing how the set point is calculated. Instead of running a calibration experiment to determine the set point, we intend to use the linear regression of the bacteria curve, calculate the slope of the linear regression, and set the set point as a slope value rather than an OD value. This could remove the need for a calibration experiment and allow the code to work across multiple sensors without running additional calibration experiments or set point adjustments.
Instructions
Our Assembly and User Instructions on Cadasio are largely complete but we hope to expand upon them. We envision the instruction as a place where the user can find all the information to assemble the system, so we plan on embedding the bill of materials, purchase locations for the individual components, and the corresponding data sheets. We hope to model the wiring and electronic components further and add instructions for the calibration experiment and system cleaning. Lastly, we hope to add short videos regarding troubleshooting and to accompany slightly difficult/ambiguous steps where a visual aid may be beneficial.
Running PACE
On top of our hardware goals, a large part of our plans involve testing the system with a bacteriophage and running a full PACE experiment. Up until this point we have only been able to test the functionality of our turbidostat and its ability to maintain a set point. Once we have viable genetic constructs, testing a full PACE experiment is a critical part of our future plans.