This section consists of N.I.N.O.’s physical design and the biochemical design. The physical design lists the design requirements and physical components of the N.I.N.O. The biochemical design shows the design proposal of the micro-organism used in N.I.N.O. and a growth calculator that is able to model the reactor. This is important to obtain the necessary concentration, volumes, and feeds of the reactor.

The physical design

The physical design of N.I.N.O. depends on the design requirements listed below. These requirements ensure that N.I.N.O. can be used as intended.

Design requirements

- Costs less than 10 dollars to produce

- Pocket size

- Durable (no leakage)

- Playable for 5 days

- Feedback from micro-organism

- Easy to operate

- Of-the-shelf parts

- Open source

- Symmetrical aesthetics

The N.I.N.O. production cost cannot be higher than 10 dollars. If the cost are too high a smaller group of people can afford N.I.N.O. This would undermine the objective of letting anyone play with biotechnology. The N.I.N.O. should be pocket size. Because the microorganisms are your babies, you must be able to transport them wherever you go! I finally understand why some people prefer chihuahuas over normal sized dogs. The design should be durable in order to protect both the player and the microorganisms. Meaning that no leakage can occur when used in normal circumstances. The aim is to make the game playable for 5 days. This could change if the characterization yields unexpected results. The user should be able to receive feedback on its performance. In other words, the micro-organism should be able to show whether it is happy or not happy by communicating with the player. Using the N.I.N.O. should be easy. It should be as easy as operating a water ring toss toy. The N.I.N.O.'s ease of use results in a larger audience that can be reached. The N.I.N.O. should be built with off-the-shelf parts. This makes it cheaper to produce and it makes it easier for the community to adapt N.I.N.O. to their liking. This is also why the blueprints of N.I.N.O.'s design should be open source. Finally, the aesthetics should be symmetrical, meaning the design has no indication of how to use the device.

Design schematic

Figure 1 shows a schematic of the N.I.N.O’s design. The bioreactor is the device's main component. The microorganisms are hosted in this part, which is visible to the player. The maximum volume inside the bioreactor is greater than the initial volume. This is because the design of N.I.N.O. makes it a fed-batch system. To prevent anything growing, the initial volume only contains phosphate buffer. This is even true when the start button is pressed. The biochemical design determines the initial volume and bioreactor volume. Next to the bioreactor there are 2 more reservoirs containing different solutions. The first is the feedstock reservoir. The water-based solution inside contains a minimal medium with glucose as the carbon source, ammonia (NH₄⁺) as an N-source, and vitamins. The second reservoir contains the inoculum (the micro-organisms). Adding the biomass to the bioreactor starts the N.I.N.O. For every reservoir, a pump is installed to pump the liquid from the reservoir to the bioreactor. These pumps pump a set volume per activation. The pumps are activated by pressing the pump. The force of pressing down will push the liquid through its tubing. No electricity is used in the device. All the energy comes from kinetic energy produced by the user. The micro-organism will produce gas (CO2). This gas can be used to communicate with the player. It can be caught in a (growing) balloon or syringe. It can be designed to be the finish line. For example, if the balloon bursts, the game is finished or it can be used to transform the device (f.e., grow flagella).

Figure 1: The schematic of the N.I.N.O.

The biochemical design

The Micro-organism

The micro-organism is the heart of the project. It is what makes N.I.N.O playable. N.I.N.O. can be used with any type of microorganism. However, it is important for the micro-organism to give feedback to the user when it is not feeling well. Microorganisms can communicate with humans in a variety of ways, such as through smell or color. For this project, it was chosen to use color as a medium for communication. The following section explains how communication via color could be introduced into a micro-organism.

Using only one method of communication keeps the design simple. As a result, it is determined the micro-organism will only communicate when it is stressed. Stresses arise from high osmotic pressure, low or high pH, starvation due to a lack of nutrients, mechanical friction, and temperature. It would be difficult to engineer a sensor for all these different forms of stress. Therefore, a general stress marker called RpoS is used to produce a pigment⁶. This is done by using a RpoS sensitive promotor that upregulates the expression of a red fluorescent protein (RFP)5. RpoS reaches a very high concentration when the bacteria faces stress. The RpoS will change the expression of many genes to make the bacteria more resilient to stress⁷. It is important to note that not all bacteria have this RpoS system, especially some laboratory strains that have lost their RpoS. The loss of RpoS can be advantageous because switching the gene expressions takes an investment of energy⁸. It is therefore important to check if the strain used is not RpoS impaired.

Figure 2 shows the design of the suggested transformation into the bacteria. The transformation is relatively simple since only one gene is introduced. The first part of the insertion consists of a promotor (fig. 2A). This promotor should be upregulated by RpoS. Pdps is the promotor of the dps gene in bacteria and is known to be upregulated by the presence of RpoS⁸. The second gene is a red pigment. A good option is using a red fluorescent protein(RFP). There are many variants of RFPs. Suggested are mCherry, mStrawberry, and mPlum⁹. Finally, a terminator is needed. The exact method of the construction and transformation of this gene depends on the parts library and method used at Ginkgo.

Figure 2: The design of the inserted gene in the N.I.N.O. micro-organism.

Growth calculator

The playability of the game depends on the type of micro-organism used and the concentrations of substrate, products, and biomass over time. As well as the volumes of the different reservoirs and feeds. A growth calculator is made to calculate start concentrations and volumes. The growth calculator has a series of inputs. These inputs can be categorized into inputs depending on the micro-organism used, starting concentrations, and the N.I.N.O. design. The growth calculator is available via this link.

Micro-organism dependent inputs

The Yield (Y_x/s)

The yield is the ratio between the amount of biomass produced over the amount of substrate consumed in kgX/kgS. The value is between 0 and 1, and the higher the value, the more efficient the micro-organism is using the available substrate for growth. This is a value that depends on the micro-organism and should be calculated with an experiment.

The maximum growth rate (µ_max)

The maximum growth rate is the maximum growth the micro-organism achieves during exponential growth (no nutrient limitation in ideal circumstances). The higher this value, the shorter the doubling time of the micro-organism is. The value depends on the micro-organism and should be calculated with an experiment.

The substrate specificity constant (K_S)

The substrate specifity constant is the substrate concentration where the growth rate is exactly half of the µ_max. This is a micro-organism specific constant and should be obtained from experiments. The units of the constant is kgS/m³.

The cell specific death rate (K_d)

While organisms have a cell specific growth rate (µ), they also have a cell specific death rate. When the cell specific growth and death rate are equal, the organism enters the stationary phase.

The product yield constant acetate(Y_p,acetate/s)

In anaerobic fermentations, micro-organisms can not fully combust their substrates. Meaning that there are some catabolic products. While in general, products are not interesting for N.I.N.O., some products have an inhibitory effect on the growth rate depending on their concentration. Acetic acid/acetate is a product that is known to have an inhibitory effect on the growth rate^10,11.

The product yield constant ethanol(Y_p,ethanol/s)

Another product of the fermentation that has an inhibitory effect is ethanol¹⁵. The yield is necessary to know to calculate the effect it has on the growth rate.

The product specific inhibitory constant acetate (K_P,ace)

Acetate has an inhibitory effect on the growth of micro-organisms. The specific inhibitory constant is necessary to calculate this effect.

The product specific inhibitory constant ethanol (K_P,eth)

Ethanol has an inhibitory effect on the growth of micro-organisms. The specific inhibitory constant is necessary to calculate this effect.

The product yield constant CO₂(Y_P,CO2/S)

The volume of gas produced during the fermentation is an important design parameter. The size of the off gas container depends on it. To calculate the volume of gas produced, you need the yield of CO₂ over substrate.

Starting concentrations

Substrate concentration at t=0 (C_S0)

The starting concentration of substrate is important for the growth curve and rate of the organism. If the concentration is very high, the organism can grow at its maximum growth rate for a longer time and additional feeding is not necessary. If it is zero, the micro-organisms will not grow unless feed containing substrate is added.

Substrate concentration in the feed (C_s,in)

The substrate concentration in the feed together with the feed volume determines the amount of substrate entering the reactor when the feed button is activated.

Biomass concentration at t=0 (C_X0)

The initial biomass concentration is important, because a high concentration will cause the game to end earlier. This amount is therefore dependent on the designed t_final (max total time of the game).

Starting volume in the bioreactor (V₀)

The starting volume of the bioreactor is important because it determines the left-over volume to be fed into the reactor.

Activation interval (h)

The activation interval is the time between two activations of the feed button. The shorter the interval the more the reactor is fed. Feeding every 10 hours might be a bit boring, while feeding every 5 minutes is too much.

N.I.N.O. design dependent inputs

The maximum bioreactor volume (V_max)

The maximum volume of the bioreactor is the limited factor of the fed-batch principle of the N.I.N.O. bioreactor.

Feed volume per activation (F_act)

The amount of added volume when a button/pump is activated.

Calculations performed by the model

These inputs are used to calculate the further course of the reactor. With this information, you can design the N.I.N.O in a way that it is playable. The most important parameters are the concentration of biomass, the concentration of substrate (glucose), the growth rate, the concentration of ethanol, and the pH.

Euler steps model explanation

The calculations described below consist mainly of derivatives. The easiest way to solve these derivatives is to use the Euler’s method or Euler steps. It is a numerical method to approximate the actual outcome. The Euler method takes small time steps and evaluates the direction of the point in the graph. In this case, the X-axis is the time and the Y-axis the concentration of a certain compound. Subsequently, the direction is the rate of change of concentration of this compound. When you take this direction/slope at a time point X, you can approximate the Y value at X+1. The smaller steps you take, the more accurate the model calculates. In this case, we use a step of 1 second. On a whole day (86400 seconds), these steps are quite small. Therefore, the approximation is also quite good.

The growth rate (µ) is dependent on the maximum growth rate (µ_max), the concentration of substrate (C_S and K_S, blue part of the equation), the concentration of acetate (C_P,ace and K_P,ace, red part of the equation), the concentration of ethanol (C_P,eth and K_P,eth, pink part of the equation). The formula is depicted below this paragraph. The growth rate is important for project N.I.N.O. because a low growth rate for a longer period of time will cause the bacteria to produce the red dye. If the calculations show that with the design of the bioreactor it is impossible to keep the µ > 0, the design should be altered because the game is deemed unplayable.

The actual rate of change in biomass due to growth depends not only on the growth rate but also the concentration of biomass available at that point in time (C_X). This rate is later used to determine the amount of biomass at a certain time point.

Next to the growth rate of the bacteria, there is also a death rate (K_d) that is always there. The actual change of biomass due to this deathrate depends also on the available biomass at a certain time point (C_X).

Subsequently, the actual rate at which biomass changes (r_X) is the change due to growth minus the change due to death.

The change in substrate (r_S) is calculate using the yield of biomass on substrate (Y_X/S) and the rate of biomass growth (r_X,growth), since the energy of the substrate is used to grow biomass.

Because the growth is anaerobic, the sugar can not be combusted, but is changed into a variety of other catabolic products. Two of these products, acetic acid and ethanol, are interesting because they have an inhibitory effect on the growth of the micro-organism. The next two equation show the calculation of the rate, which is based on the yield of acetate on the substrate and likewise for ethanol.

Another catabolic product of interest is CO₂. This is a gas and its volume is important to know because it determines the size of the off gass reservoir. Because it is a catabolic product the rate of production is depend on the substrate rate and the yield of CO2 over the substrate.

Now that all rates are defined it is possible to calculate the mass balances over time for the biomass, the substrate, and three two products. First the substrate, the amount of substrate over time in the bioreactor is depending on two parts, the amount of substrate entering the reactor at that moment and the consumption of the substrate already in the reactor. This yields the following equation:

Next up is the biomass, this mass balance also consists of the two parts, the increase of biomass and the income concentration of biomass. Since, the inflow does not contain any biomass a flow pulse decreases the concentration of biomass.

The mass balances of acetate and ethanol are produced in a similar way because they as well increase over time.

Because CO₂ is a gas and the assumption is made that it is not dissolving in the solution. The mass balance is equal to the rate of CO₂ production.

To calculate the amount of off gas produced Henry’s law is used. Henry’s law states that the concentration of a gas in a solution is proportional to the partial pressure of this gas and the concentration in the liquid. This proportion is summarized by the henry’s constant (K_H).

The K_H of CO₂ is 0.037 mol/L*atmosphere. It is assumed that only CO₂ is present and that the N.I.N.O. is under atmospheric pressure. This means that the partial pressure is 1 atmosphere. This means that the concentration of CO₂ in the bioreactor needs to be >0.61 g/l before off gas is produced.

Showcasing the model

The goal of this model is to make a N.I.N.O that is playable. The first step is to characterize the micro-organism used. In this case the values used will come from the characterization of W3110 e. coli ‘ Wildtype’ strain¹³ .Table 1 shows the parameters that are used as input for the growth calculator. The italic parameters are easy adjustable, the non-talic parameters are inherent to the used micro-organism and its environment (temperature, type of medium, etc.).

Table 1: Key input parameters of the growth calculator and their values.

*Values come from table 3 and 4 of REFXXX

** Value is not in the report and is made up

When the values of are used as an input for the growth calculator the following graphs are returned (see figure 3). The top left panel shows the concentrations of glucose, biomass, acetate, and ethanol. At first the concentration of glucose is increasing because the rate of consumption is lower than the amount of glucose added by the feed. After approximately 6 hours this starts to change and consumption goes faster than the growth. This has to do with the increased biomass concentration, more cells means faster consumption of glucose. After 12 hours all glucose gets consumed before new glucose is fed into the reactor. This immediately has an effect on the growth rate (bottom left panel). At first the growth rate sky rockets to 80% of the maximum. When the glucose runs out the growth rate drops to zero until new glucose is fed into the reactor. The effect of this is also seen in the biomass concentration. It shows that the population is going into a stationary phase (plateauing). This is something that will increase the RpoS concentration in the cells which leads to the production of RFP.

Figure 3: Resulting graphs after running the growth calculator with the inputs shown in table 1.

Using this information you can deem the game unplayable. The next thing you do is turn some knobs of the design to see whether you can make the game playable. With a maximum growth rate of 0.44 it is probably difficult to make a game last 72 hours without getting into stationairy phase. So we limit ourselves to 1.5 days (36 hours). We will also increase the amount of sugar in the feed by increasing the feed volume per activation from 0.5 to 1ml, and decreasing the start volume to 10ml. This yields the following figure.

Figure 4: Results of the growth calculator after altering some design parameters.

Compared to figure 3, figure 4 has a longer period of exponential growth and its biomass graph is not plateauing. Combined with a larger feeding interval at the start and a smaller at the end you can argue that this game is playable. When a game is deemed playable you can now extract the design parameters from the calculator. In this example, the size of the offgas reservoir should be 3.5ml, the size of the feed reservoir should be 18 ml, the size of the reactor 28ml, etc.

Model assumptions and limitations

The model is not perfect. This is because there are some assumptions made and the model has its limitations as well. This paragraph explains the assumptions made and the known limitations of the bioreactor. However, it could be that there are some limitations that are unknown that decrease the accuracy of the model. In this case it is important to find out what is causing the drop in accuracy and try to solve it empirical.

The first assumption is that the pH is buffered well enough that the pH does not drop significantly because of the CO₂ and acetic acid in the solution. A too low pH has big consequences for the growth and the stress levels of the micro-organism as well as the solvability of CO₂.

The second assumption is that the maintenance of the micro-organism does not change. A rise in its maintenance (the energy needed to keep a cell running) can cause a drop in yield of biomass over substrate. In this model the yield is set at a certain value. The maintenance can rise because of certain inhibitors (like acetate).

Concept design

Figure 5 shows a concept design. In the first sketch, the concept design is depicted. It shows a symmetrical design. The reason for the symmetrical design is that it can be used in any way the user wants. The N.I.N.O. has two buttons: the start button to inoculate the reactor (green) and the feed button (blue). Both buttons will add liquid to the main bioreactor, increasing the volume inside (image 2). When the micro-organisms grow, they will start to provide feedback. If they are stressed, a red pigment (RFP) will be produced as shown in image 3. The fermentation of glucose (the substrate used) will result in the production of gas (CO₂). This gas will inflate the flagella, thereby communicating to the player that they are alive and growing!

Figure 6 shows the time-line of the project. The total project consists of multiple chapters. These chapters are.

- Transformation of the micro-organism

- Characterization of the micro-organism

- Biochemical process design (using the growth calculator)

- The design of the N.I.N.O

- Testing cycle

- Project delivery

Preparation before the start of the project

Preparations are necessary to make sure the project is off to a good start. These preparations include discussing the plan with experts to reduce the number of mistakes and do-overs. Reserving project requirements such as lab space. Ordering customized lab consumables such as DNA parts, primers, etc. Fixing all the hassle around the stay, f.e. a visa, a place to stay, etc.

Transformation of the micro organism

The first item on the to-do list is the selection [dv1] and transformation of the micro-organism. When selecting for an organism, the characteristics that are important should be considered. The strain should be easy to modify/transform, RpoS-expressing, and the maximum growth rate in an anaerobic environment should be in the right range. With the transformation, the red dye producing function is built into the genome of the organism. This red dye provides feedback to the player.

Characterization of the micro-organism

After the micro organism is chosen and the RFP gene is transformed into the organism, the strain needs to be characterized. Therefore, a batch fermentation is run which gives the necessary constants to be used in the growth calculator. These constants are µ_max, K_S, K_d, Y_X/S, Y_P,acetate/S, and Y_P,ethanol/S. The second purpose of the characterization is to see under what conditions and how much the RFP is expressed.

N.I.N.O. design and prototyping

After the micro-organism is characterized, the design and prototyping of the actual reactor can start. The physical design focusses on the usability and aesthetics of N.I.N.O. The final design should meet the requirements set out in the design section. The design of N.I.N.O. will be done parallel to the biochemical design (next paragraph).

Biochemical design

The biochemical design is made using the growth calculator. With the input from the characterization, the growth calculator is used to obtain the biochemical design. The biochemical design consists of the max volume of the reactor, volume of the feed reservoirs, feed volume per activation, start concentrations, and concentrations of chemicals in the feed.

Testing and improving

While the growth calculator and the characterization will mitigate the risk of failure in design, the system should be tested and tweaked to ensure that the game is playable. This is done in the testing and improving phase.

Project delivery

The testing and improving phase yielded a playable N.I.N.O. The deliverables of the project are made in the final week. The deliverables consist of a presentation, an N.I.N.O. prototype, and a blueprint of the N.I.N.O.

The project needs resources to achieve the ideal outcome. Naturally, these resources come at a cost. Both the necessary resources and their costs are listed in this chapter. Note that the estimations of cost can diverge from reality due to the many unknowns, like lab work flow, salaries, and consumable costs.

Table 2: Estimated costs for project N.I.N.O.