Results
Below are the results of our experiments!
Lysis Team
Dummy 1:
Rz1 is our most effective lysis gene of the three that we had prepared at the time. For all three genes, the lysis genes kept the OD600 – and thus the population – lower than in the control.
Dummy 2 :
Rz1’s response time was especially fast for 50 ng/ µL but 100 and 200 were effective too. For all three genes, the lysis genes kept the OD600 – and thus the population – lower than in the control.
Experiment 1:
As shown in the graphics, Phi X was the most effective across concentrations, while Lambda Holin and Rz/Rz1 both demonstrated cell lethality that was a little less efficient.
Experiment 2:
Because of issues with time and getting the transformations to stick, we were only able to run experiment 2 with Rz/Rza and PhiX.
As seen below, PhiX shows a clear distinction in the OD-controlled GFP fluorescence of its supernatant. Rz/Rz1, however, also demonstrated increased GFP levels compared to the control (although, relatively speaking, they are much lower).
Experiment 3:
The graphic below shows the effect on the various lysates on mammalian cells. The bars show the proportions of mammalian cells alive versus dead after 5 hours. The green dots show the average % of the samples that were alive. When the lysate was added, overall population decreased (even with the control), but the average percentage of the cells alive was unchanged.
Experiment 4:
The graphic below shows the effect of lysate on a population of mCherry expressing, non-lytic E. coli which act as a model of other bacteria in the stomach that will be exposed to the lysis genes via the BOOMcoli lysate. As the graph shows, PhiX left the highest amount of the mCherry bacteria alive compared to the LB control.
Quorum Sensing Team
As illustrated in Fig. 1, the engineered bacteria can produce sfGFP as a result of quorum sensing, when they reach a certain optical density. The production curve of sfGFP provide reference for the dynamics of quorum sensing over time. By taking the initial fluorescent intensity in Fig. 1A, we identified the point in time where the fluorescent intensity increased by 50%. The optical densities (Fig. 1B) at that point in time are displayed in Table. 1 as it shows the relative threshold OD for the quorum sensing bacteria. The threshold ODs for different initial dilutions are similar when accounting the standard deviation.
Figure 1. Engineered E. coli produces sfGFP at high cell density. (A) represent the fluorescent intensity measurements. (B) represent the optical density measurements. Overnight cultures were diluted with different initial ratios and measured on plate readers. The legend indicates the initial dilution ratios. Each dilution was done in four replicates. The shaded areas represent the standard deviation.
Table 1. Threshold OD for quorum sensing.
We then integrated lysis protein production and quorum sensing by replacing sfGFP with the gene for lysis protein Rx production. To assess whether our integration worked, we back diluted our engineered bacteria that produces lysis protein, Rx, via quorum sensing and a control strain. Similar growth was initially observed between the experimental and control group until approximately 300 minutes. Then, the growth curve of the experimental group plateaued while the control group continued growing, showing that the lysis protein has been produced and are effective in the experimental group.
Figure 2. The optical density of the engineered bacteria that produces lysis protein, Rx, via quorum sensing (yellow) and of a control strain that does not produce lysis protein (black) are measured. Overnight cultures were diluted with different initial ratios and measured on plate readers. Three replicates were done for each group. The shaded areas represent the standard deviation.
Future work
Based on the modeling results, the next step of this project is to integrate a calprotectin sensing promoter by replacing the LuxR promoter with it. By doing so, LuxR will be produced in a low zinc environment, which is correlated with high concentrations of calprotectin, an IBD biomarker for inflammation. Therefore, an AND gate will be produced where lysis protein is only produced at inflammatory sites when there is sufficient bacteria density to release effective dosage of therapeutics.
Therapeutics Team
Confirming the production of IL-10:
Western blot experiments were conducted on E.coli stellar cells containing plasmids Tp002, Tp003 and Tp004 to confirm the production of IL-10. Multiple rounds of western blot experiments were conducted over the summer, with the third round yielding a successful result.
Experiment Round 1:
OD600 before cell lysis- 0.5
The desired band of size ~20kDa was only observed in lane 3, corresponding to the cells containing Tp003 plasmid (TorA signal sequence). The control antibody, GAPDH was not used this round.
Experiment Round 2:
We ran another western Blot experiment on E.coli stellar cells containing the same plasmids from the previous round.
OD600 before cell lysis- 0.7
The control antibody- GAPDH, was mixed with the primary antibody solution, and the staining time was 1 hour. A band around 37 kDa was observed for the control- GAPDH; however, no IL-10 bands were seen.
We noticed that there were stop codons in all the signal peptide sequences and these lacked a start codon, so we once again repeated molecular cloning to correct the sequence. It was unexpected that we did not see any bands as we observed an IL-10 band for TP003 in the previous western blot experiment. It is possible that a mutation could have occurred that allowed IL-10 production in the cell containing TP003 plasmid.
Experiment Round 3:
A final western blot experiment was conducted on E.coli cells containing the corrected signal peptide sequence- PhoA and DsbA. Experiments were not performed on cells containing TorA because the molecular cloning performed to correct the signal peptide sequences did not work.
OD600 before cell lysis- 0.7
Desired IL-10 bands around 20 kDa were observed in lanes 5,6,8, and 10, confirming that engineered E.coli strains containing the PhoA and DsbA signal sequences can produce IL-10.
Testing the periplasmic signal sequences:
Engineered E.coli strains containing plasmids Tp006, Tp007 and Tp008 were observed under a Confocal Fluorescence microscope. The reporter gene- mCherry, was tagged to the fusion protein containing the signal peptide, and imaging was done to determine the localization of this fusion protein. Two rounds of Imaging experiments were performed, and both yielded similar results.
Imaging Round 1:
We expected to see a bright fluorescent ring around the cell, but instead observed that the entire cell was fluorescent. From the sequencing results and our analysis of the gene sequences on the plasmid, we found stop codons in the signal peptide sequence and a lack of a start codon directly before this sequence. Additionally, the mCherry reporter gene had its own start codon.
We hypothesized that the presence of stop codons in the signal peptide resulted in the formation of a non-functional protein preventing it from localizing in the periplasm.
Imaging Round 1:
Engineered E.coli strains containing the corrected IL-10 and signal peptide sequences (with no stop codons) were imaged and the entire cell appeared bright under the microscope like the previous experiment.
We think one reason for not observing a bright ring around the cell is because the mCherry reporter, present at the end of the fusion protein sequence, had its own start codon. This resulted in the mCherry reporter being expressed irrespective of the signal peptide. Due to time constraints, we were unable to do another round of molecular cloning to remove the start codon before mCherry.
Lysis + Therapeutics
Experiment 1- Il-10 Western blot
Goal: Confirm release of IL-10 via Self-Lysis.
Western Blot experiment was conducted on cells post cell self lyses containing the PhoA_IL-10 and DsbA_IL10.
Western blot experiment was successfully performed as there are IL-10 bands for 8 of 9 cell lines, showing that once the cell self lysis, it will release our desired therapeutic product.
Quorum Sensing + Lysis
We then integrated lysis protein production and quorum sensing by replacing sfGFP with the gene for lysis protein Rx production. To assess whether our integration worked, we back diluted our engineered bacteria that produces lysis protein, Rx, via quorum sensing and a control strain. Similar growth was initially observed between the experimental and control group until approximately 300 minutes. Then, the growth curve of the experimental group plateaued while the control group continued growing, showing that the lysis protein has been produced and are effective in the experimental group.
Figure 1. The optical density of the engineered bacteria that produces lysis protein, Rx, via quorum sensing (yellow) and of a control strain that does not produce lysis protein (black) are measured. Overnight cultures were diluted with different initial ratios and measured on plate readers. Three replicates were done for each group. The shaded areas represent the standard deviation.
Figure 2. Modeling results for 100 cells for quorum sensing induced lysis regardless of calprotectin concentration. The concentration of free floating AHL outside of cells and the concentration of AHL inside of cells consistently rose. LuxI, LuxR:AHL, and lysis protein concentration all rose but plateaued early on. The curves are similar to the curves created when mapping the average sfGFP produced.
Figure 3. Modeling results for 100 cells for calprotein dependent lysis under low and high calprotectin concentrations. Under low calprotectin concentration there is low concentration of lysis protein production overtime. However, for high calprotectin concentration, the concentration of the lysis protein is 20x higher.
This model was created using the following equation and variables and parameters:
