Team:Technion-Israel/Demonstrate
Introduction
Our "BeeFree" system may revolutionize the industry of honey substitutes, introducing a sustainable new product mimicking the process that is taking place in the bee's stomach. Even though we are facing a great challenge ahead that will require extensive wet-lab work to optimize our system,every component of our system has been tested and proved to be functioning independently. To reach satisfactory standards, adjusting and assembling of all components is crucial.
Circuit and Inserted Proteins
Secretion and Activity of Our Enzymes
The three main enzymes participating in the natural honey production process are invertase, glucose oxidase, and catalase, each of them is crucial to the process. Thus, the secretion and functionality of these enzymes are at the very core of our system.
We have used various methods to check for enzyme secretion and to validate that we succeeded with the secretion of the enzymes to the bacterial growth solution. Each of the inserted enzymes was secreted in its active form, as can be seen in the results chapter.
Promoter Validation Assay
Glucose oxidase (GOx) is a key enzyme in honey production. However, its activity results in increased hydrogen peroxide and gluconic acid production. These two molecules can be harmful to both bacteria and humans in high concentrations. Furthermore, the GOx increased activity would also result in an undesired final sugar concentration in the honey. Thus, GOx production must be regulated. We chose KatA as a regulating promoter, which was designed to inhibit GOx production in response to high levels of hydrogen peroxide. We have tested the promoter in increasing concentrations of hydrogen peroxide, and it appears that the promoter is, indeed, sensitive to the presence of hydrogen peroxide.
The KatA promoter has been chosen to regulate the production of glucose oxidase (GOx) in our “Honey Circuit” plasmid. As a reminder, our “Honey Circuit” plasmid is designed to secrete GOx in a regulated manner – GOx expression is regulated by the Lac promoter, and its inhibitor, LacI, is regulated by the KatA promoter. The pKatA repressor, PerR, can only bind DNA in the absence of hydrogen peroxide, and therefore only allows the expression of the downstream gene as hydrogen peroxide concentration rises. The byproduct of GOx activity is hydrogen peroxide, which enables the transcription of the gene downstream to the KatA promotor in our circuit, namely LacI, which then inhibits the transcription of GOx. Thus, a negative feedback circuit regulates GOx production.
Figure 1: The "Honey Circuit" plasmid
To test the promoter, we have constructed a vector containing pKatA, and mCherry as a reporter gene. This vector was constructed using Gibson assembly with the pBE-S commercial plasmid (TaKaRa) that served as a backbone (figure 2) and eventually transformed into Bacillus subtilis (RIK1285). The bacteria were cultured in several Erlenmeyer flasks with BioAssay (BA) medium (containing 5% LB), until it reached OD600 of 0.4.
Then, we added different concentrations of hydrogen peroxide to each flask and incubated them for another hour. The H2O2 range concentration was determined according to previous studies [1], which can be later used to test the KatA promoter of our "Honey Circuit". We measured the fluorescence intensity (612 nm) of each sample using plate-reader, and the results are shown in Figure 3:
Figure 2: mCherry expression under pKatA
Figure 3: Fold change of normalized mCherry fluorescence intensity in different hydrogen peroxide concentration – 0, 50 and 100 μM. The fluorescence intensity obtained from the medium and the wildtype bacteria was subtracted.
Our results show that in the absence of hydrogen peroxide, high fluorescence intensity measured. Meaning the promoter is leaky, although there is a significant increase in the presence of 50 μM hydrogen peroxide.
We believe that the decrease in the 100 μM concentration of hydrogen peroxide may result from protein oxidation, for some fluorescent proteins function as antioxidants, as previously shown in studies.
Honey Circuit
After the effect of the KatA promoter had been tested, we conducted an experiment aiming to test the whole genetic circuit under different conditions and over different time scales. The results indicate that our "Honey Circuit" does result in alternating repression of GOx expression over time, as desired.
We planned an experiment in order to validate our “Honey Circuit” plasmid. We aimed to examine the behavior of “The Honey Circuit” over time, in the presence of glucose and under different concentrations of hydrogen peroxide, to determine whether the glucose oxidase (GOx) production is being regulated and reaches a steady state.
Figure 4: The "Honey Circuit" plasmid
We constructed the “Honey Circuit” plasmid on a pBE-S commercial plasmid (TaKaRa) backbone and transformed it into B. subtilis. The bacteria were cultured in several Erlenmeyer flasks with LB medium, until the culture reached OD600 of 0.6. The following solutions were added to the flasks as detailed:
| 0 μM H2O2 | 50 μM H2O2 | 100 μM H2O2 | 16% glucose | Wild type (WT) bacteria |
We extracted the bacterial supernatant at five different time-points – 0, 1, 2, 5 hours, and O/N (all based on our modeling) and preformed GOx activity assay on each supernatant.
We hypothesized that GOx enzyme expression exhibits an under-damped oscillating behavior (the system oscillates with the amplitude gradually decreasing to zero): At first, we will see an oscillating concentration of GOx production, and at a certain point, the oscillations’ amplitude will decrease and reach a steady state.
This behavior can be explained by our "Honey Circuit" – during its reaction, GOx produces hydrogen peroxide, which activates the KatA promoter and results in LacI production. Consequently, the increase in LacI leads to a decrease in GOx production. As a result, the catalase enzyme that is naturally expressed in B. subtilis breaking the hydrogen peroxide down, the remaining LacI proteins are degraded, and the expression of GOx continues.
At a certain point, the system will reach a steady state in which the levels of GOx, hydrogen peroxide, and LacI will remain balanced. Our results are shown in figure 5:
Figure 5: The "Honey Circuit" Assay Results
Our results show a behavior that fits our expectations: there is an alternating activity of GOx that resembles an oscillation behavior and does not reach a steady state during the 26 hours in which we measured. We assume that performing additional measurements will lead to a clearer oscillation display and to the discovery of the exact point when steady state is reached.
As seen in Figure 5, the starting point of each sample is different. A possible explanation is that the primary production of GOx is happening immediately after the initial exposure to the added materials. With the addition of 16% glucose, the highest GOx production has been observed, less than at 50 μM, 100 μM, and no H2O2, respectively.
When no H2O2 was added (0 μM, red graph ), GOx activity levels reached zero quickly, probably since the bacteria’s GOx expression was already at its sinusoidal decrease at t=0 (while other bacteria were at its peak). After two hours at zero level of GOx activity, the enzyme activity raised to a final value of ~0.005 [U/ml∙OD600]. We assume that after a longer period, GOx activity will reach a constant H2O2 level.
At 50 μM H2O2 ( green graph ), as well as at 100 μM H2O2 (black graph), we can see a drastic decrease in the first two hours, followed by an increase after five hours of incubation to the activity of ~0.015 [U/ml∙OD600] and ~0.006 [U/ml∙OD600], in accordance.
Finally, another decrease to a level of ~0.003 [U/ml∙OD600] was for both of the H2O2 concentrations added after 26 hours of incubation. These results can imply that H2O2 initial concentration activated the repression at the first two hours by the "Honey Circuit" as expected, and after 26 hours reached similar values of GOx activity.
At 16% glucose addition in the LB solution, similar behavior can be seen( blue graph ). The main difference is at the start and end points, which are considerably higher than other samples, reaching to GOx activity values of ~0.04 [U/ml∙OD600] and ~0.0145 [U/ml∙OD600], respectively. The higher GOx activity can be explained by the excess of GOx's substrate, glucose, comparing to the remaining samples that contain LB medium only (small amount of glucose). We assume that testing additional time-points of 16% glucose+LB samples, GOx activity value would decrease to ~0.003 [U/ml∙OD600] activity.
In conclusion, this assay supports that our "Honey Circuit" is responding to altered hydrogen peroxide levels in the solution, to a final approximal GOx activity value of ~0.003 [U/ml∙OD600].
Sugars degradation
After reviewing several studies regarding the bio-chemical process of honey production ( see design), we proved these enzymes can sustain the required process in high-volume solutions, and in our system, as designed.
The goal of this experiment is to obtain a solution comprised of the honey sugar composition using commercial enzymes, trying to mimic the process performed in the bee's honey stomach without using synthetic biology.
To follow the enzymatic process, we initially used a 10% (w/v) sucrose solution dissolved in DW and measured the pH levels over time. The temperature was maintained by placing the glasses containing the sugar solution in a heating bath fixed to 32°C. The pH measurement took place simultaneously, and the experiment was performed in triplicate.
The experiment included 3 test glasses, in each glass, a constant quantity of enzymes was added. 232, 105, and 50 units of invertase, glucose oxidase (GOx), and catalase were added, respectively. Those amounts were chosen after calculations considering the enzyme activity and the time that it takes for them to reach the desired concentrations (as detailed in the protocol).
The following samples were assembled:
Glass 1 – Invertase only. This test was made to determine the Invertase activity alone.
Glass 2– Invertase and GOx. This test was made to determine GOx influence on Invertase activity.
Glass 3 – Invertase, GOx, and catalase. The full mixture of the enzymes that will be used in our system.
Figure 6: pH changes over time for sucrose solution incubation with different enzymes.
As can be seen in Figure 6, the results of the pH measurements are as expected. Samples with only invertase show a slight decrease in pH level due to the production of glucose and fructose that are considered weak acids [2]. Samples containing GOx show a significant decrease in pH levels, an outcome of the gluconic acid production in the solution. As seen after 10 hours, the decrease in pH is moderate, as the low pH affects GOx activity.
The next step of the experiment is measuring the amount of glucose and fructose using the reducing sugars protocol. Briefly, we added picric acid which reacts with glucose and fructose to form a color reaction. Then, we measured the absorbance as shown in Figure 7 and 8:
Figure 7: Calibration curve used for the calculations of each sample reading
Figure 8: Change in glucose and fructose concentration throughout time.
In Figure 8, we can see the sucrose hydrolysis rate, which tracks the invertase activity. In the first 3 hours of the experiment, hydrolysis rates were similar between the different tests, yet as time passed, the hydrolysis slows down for all tested samples. The hydrolysis rate of the tests including GOx decreases more rapidly (green and orange series) compared to the test with invertase alone (blue), probably because of the low pH that inhibits the invertase activity.
After 24 hours, the hydrolysis rates in all the samples decreased and reached a plateau, implying a stopping trend. We assume that over time, the hydrolyzed sugar solutions reach thermodynamic equilibrium, thus leading to inhibition of invertase activity. Another option we considered is the degradation of invertase over time.
Figure 9: Changes of sugars in the full mixture solution during time.
As detailed in the protocol, we aimed to reach approximately 2000 micromole of sucrose in the final solution. As seen in Figure 9, we did not get the expected amount, and in fact, less than half of the sucrose was hydrolyzed, and the rate of the reaction decreased over time.
In conclusion, many factors influence the enzyme’s activity, such as the volume of the solution, pH, concentration of each enzyme, temperature, and presence of other organic materials. Changing each variable and revealing the exact combination to obtain the desired concentrations of honey will take a long time with no promise of success. Therefore, we designed our Honey Circuit, using B .subtilis to secrete the relevant enzymes, which regulates the sugar composition and pH levels in the honey solution.
The substrate we have chosen for our project is commercial nectar, kindly provided by Tsuf Globus. To verify that our system is suitable to process this nectar, we performed an experiment validating the activity of the enzymatic reactions to degrade the nectar's sugars. We have crossed data from earlier experiments (with our self-made sugar solutions) and our model, to calculate the necessary concentrations of commercial enzymes required for the process. We used commercial enzymes, and by checking the decrease in pH levels and the concentration of hydrogen peroxidase, we validated the activity of our system in the commercial nectar.
Figure 10: Change in hydrogen peroxide concentration over time. Detected by absorbance at 416nm, using commercial enzymes to process nectar over 20 hours of incubation at 37°C
Although the results did not demonstrate a linear increase, we assume that some of the hydrogen peroxide was decomposed due to catalase activity. We assume that after the instability shown until the tenth hour , the increasing shown is likely to continue until reaching a steady state. Thus, we have proved the enzymes ability to function in a nectar solution
Figure 11: Increasing acidity in nectar solution. Commercial enzymes were used to process nectar over 24 hours and incubated at 37°C
The decrease of the pH level demonstrated above indicates the activity of glucose oxidase.
System's Capacity and Adaptivity
Our BeeFree production system is complex and requires many adjustments and verification. We have answered the basic design requirements and tested the adaptivity to the altering conditions and capacity of several components. Thus, we proved our basic infrastructure's suitability.
This experiment demonstrates the survival of B. subtilis bacteria in a range of hydrogen peroxide concentrations: 0, 50, 100, 150, 200, and 500 μM of H2O2. The H2O2 range concentration had determined after reviewing previous studies [1]. Two of the concentration (50, 100 μM) were later used to test the KatA promoter of our "Honey Circuit". The assay has been performed by measuring the absorbance at OD600.
Figure 12: Bacillus subtilis survival at different H2O2 levels indicated by the measurement at 600 nm. WT bacteria growth in LB, 50μM of hydrogen peroxide, 100μM of hydrogen peroxide, 150μM of hydrogen peroxide, 200μM of hydrogen peroxide, 500μM of hydrogen peroxide, 0 μM of hydrogen peroxide (Control). All incubated for 5 hours at 37°C
As can be seen in Figure 12, the bacteria grew at the same rate in both lag and log phases at all of the given hydrogen peroxide concentrations, including the control that tested Bacillus subtilis growth without H2O2 addition.
System's resistance to decreasing pH
During the process of honey in the bee stomach, a key step is the oxidation of glucose to D-glucono-lactone, which hydrolases to gluconic acid, resulting in a decreased pH level.
Since the enzymatic activity may be influenced by the acidity of the solution, we have conducted experiments to determine whether our enzymes remain active in the relevant range of pH levels. In these experiments, we measured the enzymatic activity of each enzyme at four different pH levels, covering the range in which the honey-like solution can vary. The pH levels that have been tested are 4.5,5, 6, and 7. These experimental results were incorporated into the model , improving its credibility.
As for the bacterial resistance for decreasing pH levels, previous studies have shown resistance at the pH range of 4 to 7 [4].
"Bio-Castle" Capsules' Suitability
Our "BeeFree" honey production is designed to generate inside the "BioCastle" capsules, to which our bacteria will be injected. The capsules are designed to prevent bacterial leakage, while at the same time enable protein secretion.
Since those functions are crucial to our procedure, we have conducted an experiment aimed to demonstrate the capsules' ability to provide those two functions. The results indicate that, indeed, the capsules can secrete the desired enzymes in their functional form and prevent bacterial leakage. Thus, we have proven that these capsules can function well in our system.
To test protein secretion, we compared the catalase enzyme's activity in three different types of samples, performed in triplicate. These samples were: Wild-type (WT) bacteria without a capsule, WT bacteria in capsules with distilled water as an outside solvent, and a capsule containing WT bacteria with a sugar solution as the outside solvent.
The control sample contained a capsule injected solely with LB. The sugar solution samples were tested to determine whether the osmotic pressure resulting from the presence of sugars would affect protein secretion. The catalase enzyme was chosen to be our model enzyme due to its high activity (see activity results ).
After incubation of the capsules in the different solutions, we concentrated the solutions using Amicons (manufactured by Mercury). As catalase is a naturally produced enzymes in WT B. subtilis, there is an abundant amount of it in the supernatant, and therefore, additional concentration is not necessary. The average activity measured in the non-capsule samples was 12.27 U/ml, while the average activity measured in the capsule-DW samples was 8.54 U/ml, and the measured activity in the capsule-sugar solution samples was 1.7 U/ml. No activity was detected in the control sample.
Fortunately, it appears that the capsules can secrete the proteins and keeping it in an active state. As to the sugar solution sample, it appears that protein secretion is interrupted under the condition of osmotic pressure, and while the viscosity and density resulted in decreased diffusion rate, it was not entirely eliminated.
That crucial information could guide us in the future optimization of our system, utilizing a less concentrated solution that would result in decreased osmotic pressure, or using a device that would actively move the capsules, to increase diffusion rate. It is worth mentioning that the concentration of sugars found in our solution is higher than in real flower nectar, and therefore, we can predict that the proteins' secretion could be higher in real nectar solution.
Figure 13: Catalase activity measured in various solutions
In this part, we aimed to show the capsules' capacity to prevent bacterial leakage. We followed the bacterial presence and growth using absorbance measurements at 600 nm in three different types of samples (in triplicate), as followed:
- A capsule injected with LB solution only, placed in LB (LB-Containing capsules).
- A capsuled injected with WT bacteria, placed in LB (Bacteria-containing capsules).
- WT bacteria placed in LB, without capsules (Bacteria-LB).
The absorbance measured both in the bacteria-containing capsule and the LB-containing capsules samples remained at zero throughout the test, while the LB-Bacteria sample (without capsules) was gradually increasing. Therefore, one can claim that indeed, the capsules do prevent bacterial leakage.
Figure 14: Bacterial growth outside capsules measured by absorbance at 600 nm of various solutions: capsules containing bacteria, capsules containing LB, LB and bacteria without capsules, incubated for 5.5 hours at 37°C
As can be seen, the bacteria did not grow in neither the LB-capsules sample nor the bacterial capsules sample, while they grew in the bacteria-LB samples, proving the capsules' ability to prevent bacterial leakage.
Bacterial Growth in Nectar
This test was used to determine whether B. subtilis bacteria will be able to grow in the nectar solution independently, as we were aiming to demonstrate our systems' capacity to sustain itself. The results indicate that B. subtilis can grow in a nectar solution, though that growth is limited. Thus, the nectar solution provides nutrients for the bacteria while preventing them from reaching the exponential phase. Thus, we have shown our system's sustainability, i.e. we would not have to supply the bacteria with additional nutrients.
We have tested the growth of various B. subtilis strains, as followed:
- Genetically engineered bacteria inserted with invertase gene, placed in nectar solution.
- Genetically engineered bacteria inserted with invertase gene and genetically engineered bacteria inserted with glucose oxidase gene, placed in nectar solution.
- WT bacteria placed in nectar solution.
- WT bacteria placed in LB solution (Control).
Figure 15: B. subtilis growth in nectar solution over time. Bacteria expressing invertase enzyme growth in nectar, Bacteria expressing GOx enzyme growth in nectar, WT bacteria growth in nectar, WT bacteria growth in LB (control)
System Demonstration
Honey Circuit Activity in Capsules
In our attempt to demonstrate the functionality of our engineered parts as a complete system, we have conducted a test, which we believe can serve as a suitable prototype for our system. We have tested the GOx activity, secreted by bacteria containing the "Honey-Circuit" and placed in capsules with WT bacterial capsules as control.
The samples were placed in distilled water and incubated overnight (12 hours) at 210 rpm, 37°C. Afterwards, they were concentrated using Amicons (Mecrury) and tested for the activity of glucose oxidase (to see method, please see results ). It appears there were low, yet detectable secretion and activity. To conclude, due to the functionality and suitability of each component, we believe that we are not far from assembling a functional system.
Figure 16: Activity of the honey circuit placed in capsules
- Palmer C V., Modi CK, Mydlarz LD. 2009. Coral fluorescent proteins as antioxidants. PLoS ONE. 4(10):e7298
- Herbig AF, Helmann JD. 2001. Roles of metal ions and hydrogen peroxide in modulating the interaction of the Bacillus subtilis PerR peroxide regulon repressor with operator DNA. Molecular Microbiology. 41(4): 849-859.
- Cédric Guignard, Laurent Jouve, Marie Béatrice Bogéat-Triboulot, Erwin Dreyer, Jean-François Hausman, Lucien Hoffmann. Analysis of carbohydrates in plants by high-performance anion-exchange chromatography coupled with electrospray mass spectrometry. Journal of Chromatography A, Volume 1085, Issue 1, 2005, Pages 137-142.
- Petrackova D, Vecer J, Svobodova J, Herman P. 2010. Long-term adaptation of Bacillus subtilis 168 to extreme pH affects chemical and physical properties of the cellular membrane. Journal of Membrane Biology 233:73–83.
