Results

Overview

We are excited to present the following results:

• Successfully cloned and fully sequenced two new BioBrick parts, a new nitrite reductase enzyme and vanadium dependent bromoperoxidase.
• Observed potential increase in nitrite metabolism for the NFRH plasmid at concentrations around 30 mM NaNO2 with respect to the empty vector.
• Proved that our bromoperoxidase protein is being produced and works as expected.

Nitrite Reductase

Background

One of the potential ways to reduce methanogenesis in cattle is to divert resources away from methanogens to other microorganisms in the gut. Nitrogen-fixing bacteria present in the rumen compete with methanogens for hydrogen, using it to make ammonium which can then rapidly be converted to protein by other microorganisms. The biosynthetic route for this conversion is shown below in Equation 1:

Conversion of nitrate to nitrite is relatively rapid, whereas conversion of nitrite to ammonia is relatively much slower. This means that the amount of nitrogen fixation in the rumen must be limited in the organism, because the nitrite anion is substantially more toxic than either the nitrate or ammonium ions. However, if nitrogen-fixing bacteria contained a gene that coded for a more efficient enzyme to catalyze this conversion, this would theoretically allow for a much faster rate of conversion from nitrite to nitrate in the rumen. In turn, methanogenesis could be reduced, and protein production could increase for the animal as well.

As such, we set out to successfully insert a plasmid containing a nitrite-reductase enzyme into an E. coli chassis. Once a successfully-transformed strain was produced, we could then begin testing how it would fare in conditions similar to a cattle rumen.

Testing

The overall sequence of events to test our hypothesis was:

• Successfully introduce a plasmid containing a nitrite reductase gene into an E. coli chassis.
• Grow the transformed E. coli in aerobic culture to increase cell count. A culture containing the empty vector version of our plasmid would also be grown aerobically as a control.
• Transfer the aerobic cultures into anaerobic vials of varying concentrations and grow them overnight.
• Take OD readings on the cultures to observe survivability.
• Analyze the supernatant using Nessler’s Reagent. (Nessler’s reagent reacts with ammonia in solution to form a brown precipitate. The mass of this precipitate serves as a gauge of nitrite metabolism.)

Data Analysis

For the aerobic growth portion of the experiment, two cultures were grown for both the experimental and control bacteria. This was done so that in the event that one of the cultures did not grow for either the control or experimental groups, that there would be a backup. The optical density readings for these aerobic cultures are displayed for all four cultures, for each round of testing, below in Figure 1:

Unfortunately, the experimental cultures used in the first three rounds of testing had not been correctly transformed and did not actually contain the nitrite reductase gene. In addition, the control cultures in the first three rounds contained the red fluorescence protein, and were not true control groups. Also, the chassis used was Genehog.

However, round four yielded useful data. The control cultures contained the empty vector version of the plasmid used in the experimental. Also, the experimental plasmid was confirmed to contain the nitrite reductase gene (NFRh) by sequencing. Lastly, DH5α was used as the cell chassis instead of Genehog.

Only one of each culture (experimental and control) was used for the anaerobic portion of the experiments. The aerobic culture which yielded the higher optical density was the one used to form the anaerobic cultures. Optical densities for the anaerobic cultures were obtained after 24 hours of growth. Anaerobic data for round three was not obtained, as the experiment was halted due to the fact that the experimental cultures were found not to contain the nitrite reductase gene. Optical densities for rounds 1,2 and 4 are shown below in Figures 2-4, respectively:

Round four data indicates that, in the absence of nitrite, the control culture replicated much more quickly than the experimental. This is consistent with predictions; the additional transcription of the nitrite reductase enzyme increases cellular metabolic burden with respect to the empty vector culture. However, even at 10 mM nitrite concentration, cellular activity in the control had been depressed to be on par with the experimental, suggesting that the environment had become too toxic for the control to replicate rapidly. Performance thereafter largely flatlined in both the experimental and control groups, suggesting that the environment was too toxic for either group to replicate successfully. The experimental group grew slowly even with zero nitrite present, suggesting that the strain used simply does not replicate well in general. As always, potential error in lab procedure could have had an impact on the results. Any trends present in the first two rounds of testing are not relevant to the experiment because the experimental cultures did not contain a nitrite reductase enzyme.

Next, ammonia concentration was determined, as the ammonia present gives an indication of the activity of the nitrite reductase enzyme. Samples were extracted from each anaerobic culture and the cells were removed. Nessler’s reagent was added to the supernatant which formed a precipitate. The precipitate was dried and weighed and the final mass was reported. Mass of this precipitate is plotted vs concentration for rounds 1,2 and 4 in Figures 5-7 respectively, below:

Round 4 results parallel the OD measurements, suggesting that cell growth is closely correlated with nitrite metabolism. Additionally, the high amount of ammonia present in the control suggest that other reactions occurring in the cell are producing more ammonia than the nitrite reductase reaction across the board. Anomalous behavior in the 30 mM range could potentially be due to increased cell performance in the region allowed for by the nitrite reductase enzyme, but because there is only one round of data, drawing any conclusions about this behavior would not be scientifically sound. More replicate trials would be required to determine if this was experimental error, or if there is something going on in this region. Once again, trends present in rounds 1 and 2 don’t have any experimental significance.

Bromoperoxidase

Background

Seaweed is now being studied as a food additive because it has been shown to reduce the amount of methane produced by cattle. The compound within seaweed that is responsible for reducing methane is bromoform. Bromoform works by inhibiting the efficiency of the methyltransferase enzyme by reacting with the reduced vitamin B12 cofactor required for the second to last step of methanogenesis. We set out to successfully insert a plasmid containing a bromoperoxidase enzyme from the algae corallina pilulifera into an E. coli chassis. Once a successfully-transformed strain was produced, we could then begin testing how it would fare in conditions similar to a cattle rumen.

Testing

• Successfully introduce a plasmid containing the bromoperoxidase gene into an E. coli chassis.
• Perform protein purification procedures to isolate the bromoperoxidase enzyme.
• Use the isolated enzyme in the monochlorodimedone enzyme assay.
• Mix the isolated enzyme with potassium bromide, monochlorodimedone and the MOPS buffer in a UV Transparent plate reader.
• Record the absorbance at 290 and ensure that it is constant.
• Add hydrogen peroxide to catalyze the reaction and record the decrease in absorbance at 290.

Data Analysis

After a couple of attempts we successfully cloned bromoperoxidase into the pSB1C3 backbone. We had an inclination that the eighth colony on the bromoperoxidase agar plate was cloned correctly when we performed PCR with primers that replicated at the prefix and suffix and obtained the results shown below.

Under the B8 label two bright bands can be seen. This was a good sign that our gene insert was successfully cloned into E.coli. We used the iGEM primers VF2 and VR that were designed to replicate at the prefix and suffix site. After we obtained a thick bright band as a result we knew that the prefix and suffix sites were present in our plasmid. The next step was to sequence our plasmid to ensure that it was cloned without deletions or mutations. We spent the next week or so fully sequencing the plasmid. The sequencing results were positive and ensured us that the bromoperoxidase gene was successfully cloned into E.coli.

After receiving this great news we immediately started protein purifications to see if the bromoperoxidase protein was being produced. If the E.coli is producing the bromoperoxidase protein then a band should be seen at 64 kDa. The resulting SDS Page is shown below.

There is a small band present in the 500 mM imidazole wash at approximately 64 kDa. This indicates that a small amount of enzyme is being produced. This was a positive sign. Due to our constitutive promoter we were worried that our protein might not be expressed at a noticeable rate. We would suggest that if any other team continues this experiment they should start from the beginning with an inducible promoter. This will help later with characterizing the part. With the constitutive promoter only a small amount of the enzyme was being produced at its natural expression rate.

Next we performed the monochlorodimedone enzyme assay. We used cell lysate first and a control composed of bromide ions, monochlorodimedone and a buffer, but no cell lysate. We catalyzed both solutions with hydrogen peroxide according to the protocol.

If the monochlorodimedone is being brominated by our enzyme then the absorbance at 290 should decrease. This is exactly what we found.

The results above indicate that the enzyme is performing its designed function because monochlorodimedone is being brominated. This is shown by the decrease in absorbance in the cell lysate compared to the decrease in absorbance of the control solution. Although both the control and the cell lysate show a decrease in absorbance, the cell lysate has a steeper slope indicating greater absorbance. The data points for the cell lysate also had a higher R2 value of .889 compared to the R2 value of .195 (control). This shows that the data points from the cell lysate are significantly closer to the fitted regression line. This is a good sign that our data is following a trend and decreasing while the control seems to be increasing and decreasing randomly.

The next step in analyzing the data is to determine how many units of monochlorodimedone can be brominated per milliliter of enzyme. One unit will catalyze the conversion of 1.0 micromole of monochlorodimedone to monobromochlorodimedone per min at pH 6.4 at 25°C. To do this you use the equation shown below

After this calculation was done the cell lysate was found to be converting 0.0043 units of monochlorodimedone per milliliter of enzyme. This proves that the enzyme is working as expected. To perform further characterization more assays were performed.

Next the cell lysate containing bromoperoxidase was tested again to ensure that the results still showed a decrease in absorbance at 290. The cell lysate was compared to the control solution along with a cell lysate containing the empty vector pSB1C3.

Both reactions were catalyzed with hydrogen peroxide as referenced in the protocol.

The above results further display that our part is working as expected. When hydrogen peroxide is added to the cell lysate containing bromoperoxidase there is a negative slope indicating a decrease in absorbance When hydrogen peroxide is added to the empty vector the slope has a positive value indicating an increase in absorbance. Both cell lysates stick to their regression lines roughly the same amount. The bromoperoxidase cell lysate has a R2 value of 0.693 and the empty vector cell lysate has a R2 value of 0.652. This shows that both reactions followed their trend line to a significant degree. This is good news because it means that our data was not random, instead it was being controlled by the bromination of monochlorodimedone into monobromochlorodimedone.

In this trial the cell lysate was found to be converting 0.00094 units of monochlorodimedone per milliliter of enzyme. The cell lysate with the empty vector was converting -0.0019 units of monochlorodimedone per milliliter of enzyme. This makes sense because the empty vector was not brominating monochlorodimedone.

Next we compiled the data from all of our trials using the cell lysate with bromoperoxidase and compared it to the control solution that was used in every trial. Below is a graph showing the results.

Statistical analysis shows that the cell lysate from E. coli containing the bromoperoxidase plasmid has significant activity. Although the cell lysate showed positive results we wanted to test the purified enzyme. Unfortunately, we were not able to perform a test on the purified enzyme itself, because there were issues with the protein purification method used. We used the Ni-NTA Purification System, but in the literature the DEAE-Sepharose Fast Flow column was used. If another iGEM team were to continue our project I would suggest avoiding the Ni-NTA Purification System and using the DEAE-Sepharose Fast Flow column method. We hypothesize that the His tag purification method did not work well because our enzyme is vanadium dependent. The vanadium ions at the active center act similarly to the immobilized nickel ions present in the resin. Because of this, the nickel ion may interfere with the binding of the vanadium ion in the active site and cause the purified enzyme to lose activity.