Team:Technion-Israel/Results
Introduction
Our goal is to achieve the same composition of sugars as in bee-made honey, which will turn to the familiar substance called "honey" as soon as the solution will be dried, just like the bees do it.
We aimed to design a system that will monitor the solution conditions and regulate the production of its contents accordingly.
The chosen bacteria for that mission was Bacillus subtilis RIK1285, a bacterium known to effectively secrete proteins, while E. coli (Top 10/ 5 alpha/ stellar) used as a tool for the cloning and construction of our recombinant enzymes.
To obtain our self-regulated system, we decided to use synthetic biology tools to create our special "Honey Circuit", described in detail in the "design" chapter. For the fabrication of this circuit, we used a "pBE-S" plasmid backbone, which is a B. subtilis/ E. coli shuttle vector provided by the TaKaRa company.
The Honey Circuit we created was cloned into pBE-S backbone ligated with a signal peptide at the "aprE SP" area and inserted with gBlock at the "Multi cloning site (MCS)" area (Figure 1). The rest of the clones were created similarly.
Figure 1: Map of pBE-S vector plasmid from TaKaRa
(2018. B. subtilis Secretory Protein Expression System. TaKaRa Bio inc.)
Our target proteins are Glucose Oxidase (GOx) and invertase, enzymes that degrade glucose and sucrose, respectively, and are key enzymes in the creation of honey inside the bee's stomach.
Initially, we have chosen two known B. subtilis signal peptides, AmyE and SacB, as candidates for the secretion of our target proteins. The signal peptides employed to enable the secretion of the translated proteins which are genetically located right after them, downstream in the plasmid.
After successful insertion of the signal peptides and the proteins into the backbone by ligation and Gibson assembly, we have examined four approaches to concentrate the proteins in the supernatant and lysate: His-Tag column, TCA, Freeze-drying and Amicon Ultra-15 Centrifugal Filter . The concentration process was essential to enable us to see the proteins in the SDS-PAGE assay (30 micrograms of protein per ml are required) and determine the presence of the desired enzymes, either in the bacterial lysate or supernatant.
After additional optimization of the gBlocks, we reordered them as new sequences, not containing a stop codon between the enzyme gene and the Histag following it, as it is normally found in the pBE-S backbone. By using a Histag column, we minimized our efforts concentrating the proteins and separated them more specifically from the rest of the proteins in the supernatant or lysate.
To further confirm the presence of GOx and invertase in the solution and to determine whether they are active, we conducted protein activity assays.
Secretion
Introduction
The goals of the following experiments were to confirm the expression and secretion of our target enzymes. In order to determine the location of the enzymes, both the bacterial lysate and the supernatant were tested. First, the samples were concentrated via several methods. Then, SDS-PAGE and Western blot methods were employed in order to confirm the presence of the enzymes. Finally, activity tests have been conducted.
Method
Figure 2: Experimental scheme
Bacteria were grown in LB and separated into lysate and supernatant.
The proteins were concentrated using the following methods: TCA, Freeze-Drying, Amicon Ultra-15 Centrifugal Filter and His-Tag column.
To check our results, we used SDS-PAGE, Western blot and activity tests.
As part of our project, we inserted several combinations of signal peptides and our two proteins, Glucose oxidase (GOx) and invertase, into our model bacteria. Then we tested the protein expression and secretion.
Each test was divided into these subgroups:
Table 1: The Bacteria samples tested
2. Library – A collection of 173 different signal peptides, that were inserted simultaneously into the bacteria. Assuming at least one signal peptide is successful. Testing both the bacteria lysate and supernatant.
3. SacB signal peptide - Testing both the bacteria lysate and supernatant.
4. Honey Circuit – Our BeeFree synthetic circuit with the AmyE signal peptide and GOx enzyme. Testing both the bacteria lysate and supernatant.
5. No Signal Peptide (SP) – Bacteria with the plasmid with no signal peptide, testing only the bacteria lysate.
In addition, we used a Wild type of B. subtilis as a control native bacterium.
Results
Here, we present the progress we have made to identify the proteins for our BeeFree system using the different methods mentioned above.
Figure 3: SDS-PAGE test results using His-Tag concentration for Invertase with AmyE signal-peptide
As can be seen in Figure 3, our invertase His-Tag concentration did not result in any visible band in the expected area of invertase.
Figure 4: SDS-PAGE test results using His-Tag concentration for GOx with AmyE signal-peptide
Figure 5: SDS-PAGE test results using His-Tag concentration for GOx without signal-peptide
Figure 6: SDS-PAGE test results for the Honey circuit by His-Tag concentration method
As can be seen in figures 4-6, there is a weak band corresponding to the size of the standard GOx. However, it was difficult to reach a concrete conclusion from looking at the gels presented in figures 4-5, since we had encountered a problem with our WT samples preventing us from running them in the gel. Luckily, the WT supernatant sample did run, as can be seen in figure 6, and no band was observed.
The result for the WT supernatant was encouraging since we have detected the protein in its expected size in the "Honey Circuit" supernatant, and not in the WT sample.
Moreover, there was a strong band in the range of 40-48 kDa in the sample containing the AmyE signal peptide (Fig. 4,6), as can also be seen in the Western Blot experiment, about which we will elaborate below.
In summary, we got inconclusive results from our SDS-PAGE runs. The gels seen in figures 4-5 gave us a vague assumption of the expression and the secretion of GOx since we couldn’t compare it to the WT. The "Honey Circuit" results (figure 6) were more positive, as we detected a protein in the expected size, and could compare it to the WT.
Encountering time limitations, we decided not to repeat those tests but to move on to a more specific test- the Western Blot.
Table 2: Summary of the SDS-PAGE results
Western Blot
Western blot (WB) is a more precise method compared to SDS-PAGE. By using antibodies to detect our protein's His-tag, it can be guaranteed that our protein had been expressed and secreted by the bacteria.
Figure 7: WB test results for the Honey circuit by His-Tag concentration method
As seen in Figure 7, the WB test yield a definite result in the Honey circuit strain's supernatant, meaning that our bacteria express and secrete the GOx protein.
However, the results are somewhat unexpected. Firstly, the size of the protein was different than the expected size, which according to the standard and other studies was is in the range of 67-80 kDa.
Nevertheless, we could not find a reliable source explaining the different size of our GOx. Our hypothesis is that the signal peptide affects the translation of the enzyme in the bacteria, resulting in a different structure of the protein monomer. Another assumption we have is that a part of the protein stays in the membrane during the secretion process as part of the secretion pathway, as described in the "description" chapter, which affects the protein size.
Secondly, The Honey Circuit contains the AmyE signal peptide, yet the sample containing AmyE+GOx was not visible in the WB. The solution to that may be the difference between the promoters of each strain, the Honey Circuit strain promoter is pLac and the AmyE GOx strain promoter is pAprE.
Nonetheless, the experiment ought to be repeated in order to get a conclusive result and to optimize our system.
Enzymatic Activity Test
We have performed plenty of assays to confirm that our inserted genes lead to the production and secretion of our target proteins. However, we also need to verify that our enzymes remain active. Therefore, we had performed several assays that verified and quantified the enzymatic activity of the three enzymes created by the recombinant Bacillus subtilis: invertase, glucose oxidase and catalase.
Invertase Assay
Invertase is an enzyme that cleaves sucrose to form glucose and fructose. Thus, a precise method that will detect either glucose or fructose can indicate the amount of sucrose cleaved by the enzyme. The test we used is based on the "reducing sugars" assay, in which reducing sugars react with the yellow picric acid to form picremic acid, which is red (Figure 8). Therefore, the absorbance of red light (492nm) could indicate the amount of sucrose cleaved, which indicates the enzymatic activity.
Figure 8: Illustration of the reaction of picric acid with reducing sugars
To verify that the activity-absorbance correlation remains linear under the experiment's conditions, we performed the experiment using a commercial enzyme, creating a calibration curve. We used increasing concentrations of commercial invertase (with a well-determined specific activity) and a constant saturated concentration of sucrose. Then, we checked the 492 nm light absorbance and plotted the calibration curve (Figure 9).
Figure 9: (A) Calibration curve for invertase activity – change in absorbance at 492 nm for different activity values. (B) The calibration curve tubes
After the cloning process had been completed, we performed an activity test on the bacterial invertase. We tested both bacterial lysate and bacterial supernatant to determine whether active invertase had been produced and secreted by evaluating the activity in each sample. The recombinant bacteria samples were compared to wild type (WT) samples.
The first colony tested contained the invertase enzyme with AmyE as a signal peptide. Unfortunately, no notable activity was detected (Figure 10).
These results, in addition to the results observed in the SDS-PAGE experiment, indicating that the AmyE signal peptide may interfere with either the enzyme's production or its activity.
Figure 10: The activity of invertase and AmyE in different samples, indicated by the absorbance at 492 nm
In addition to evaluating the activity of invertase with AmyE, we used a library of signal peptides aiming to obtain enzymatic activity for the bacterial invertase (FigureX4). The Bacterial supernatant activity detected is 0.18U/ml and the activity detected in the lysate solution is 0.12U/ml. The activity detected for the WT sample is observed to be significantly lower in comparison to the bacterial samples, both in the supernatant and the lysate, i.e, one or more of the library signal peptides has enabled both secretion and activity of the enzyme.
Figure 11: The activity of invertase and Library in different samples, indicated by the absorbance at 492 nm
Glucose Oxidase Assay
Glucose Oxidase (GOx) catalyzes the oxidation of glucose to D-gluconic-lactone and hydrogen peroxide. Therefore, a precise method to determine the amount of hydrogen peroxide found in the solution could indicate the activity of the glucose oxidase. In this experiment, we used the assistance of another commercial enzyme- horseradish peroxidase (HRP) and 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) (ABTS). Using hydrogen peroxide, the HRP catalyzes the oxidation of reduced ABTS, which is colorless, to oxidized ABTS, which is turquoise. While using excess ABTS and HRP, the absorbance of 416 nm light could indicate to us the amount of hydrogen peroxide created by the glucose oxidase, and therefore - the glucose oxidase activity.
The test is based on the following reactions:
To verify that the absorbance-activity relation remains linear under the experiment's conditions, we performed the experiment using commercial GOx with increasing activity, measured the 416 nm light absorbance and plotted a graph to demonstrate that relation. Fortunately, we have proved the absorbance-activity relation linearity under the experiment's conditions (Figure 12).
Figure 12: Calibration curve for glucose oxidase activity – change in absorbance at 416 nm in different activity values
After the cloning process had been completed, we performed the activity test on both the supernatant and the lysate of genetically engineered bacteria as well as WT bacteria as control. The first colony tested contained a gene for glucose oxidase and the AmyE signal peptide. The absorbance detected in the recombinant Bacteria was significantly higher, in both supernatant and lysate (Figure 13). The evaluated activity of glucose oxidase is 0.46U/ml in the bacterial supernatant and 0.45U/ml in the lysate solution.
Figure 13: The activity of glucose oxidase and AmyE in different samples, indicated by the absorbance at 416 nm
Aiming to obtain increased activity, we used a library of signal peptides for the glucose oxidase insert, similar to the experiment performed on invertase. Unfortunately, we could not detect any notable activity (Figure 14).
Figure 14: The activity of glucose Oxidase and Library in different samples, indicated by the absorbance at 416 nm
Catalase Assay
The catalase enzyme decomposes hydrogen peroxide to water and molecular oxygen. Since water cannot function as an indicator in aqueous solution and molecular oxygen tends to evaporate quickly, the main indicator for catalase activity had to be the hydrogen peroxide. Hence, we had performed an experiment that yielded us the hydrogen peroxide-absorbance relation, and therefore the catalase activity-absorbance relation. We had decomposed hydrogen peroxide using catalase, and afterward added the hydrogen peroxide-catalase solution to a solution containing ABTS and HRP enzyme. The blank (negative control) contained the initial amount of hydrogen peroxide and no enzyme. Each sample's absorbance was subtracted from the blank sample's absorbance to evaluate the amount of hydrogen peroxide decomposed by the catalase.
The reaction on which the test is based on is:
To validate this method for testing, we had to verify the relation between the amount of hydrogen peroxide and the 416 nm light absorbance measured is linear. To do so, we have tested the reaction of constant amounts of ABTS and HRP to increasing amounts of hydrogen peroxide. After performing the test and plotting the hydrogen peroxide concentration-absorbance graph, we have deduced that this method is in fact, reliable (Figure 15).
Figure 15: Verification of Linear Ratio between Hydrogen Peroxide and Absorbance
The catalase enzyme is naturally produced and secreted by the Bacillus Subtilis Bacteria, thus, a comparison between WT and genetically engineered bacteria is not needed for that enzyme. We had plotted a calibration curve using increasing amounts of commercial catalase enzyme and used it to determine the activity of the bacterial lysate and supernatant. From that experiment, we had evaluated the bacterial lysate activity to be 22.26 U/ml and the bacterial supernatant activity to be 12.26U/ml. These results confirm that the bacteria produce the catalase and partially secretes the enzyme.
Figure 16: The activity of glucose Oxidase and Library in different samples, indicated by the absorbance at 416 nm
Conclusions
Table 3: Summary of secretion and activity assays
As presented in Table 3, the Honey Circuit supernatant is the most successful strain (as seen in the demonstration). We have achieved positive results in all categories, SDS-PAGE, Western Blot and activity tests, for that sample. The fact that we saw the secretion and activity of GOx is encouraging as it serves as the proof of concept for the project. Despite the success we had with GOx, not all of the secretion tests turned out to be successful. Due to time limits, we have decided to continue only with the AmyE signal peptide.
We are aiming to continue and research the SacB signal peptide as well as the commercial library of 173 signal peptides to find the most suitable signal peptides for our system. Moreover, tests for unexamined samples should be performed.
