Team:EPFL/Proof Of Concept
Espress'EAU - Proof of concept
Overview of the problem
Pollution of groundwater by phytosanitary products leading to tap water contamination is a growing problem in Switzerland. Small communities, as well as several actors involved in the process of water quality testing, have been severely affected by the recent cases of pesticides contamination of the water network. A low-cost and easy-to-use solution would help them tackle this issue better.
Here we aim at showing that Espress'EAU is able to measure the growth behaviour as well as the fluorescence of engineered yeast strains. This would provide its users with an indication of the presence of phytosanitary products in the water they want to test. Thus, it could serve as a trustworthy system for generic water quality testing.
MSN2 deletion strains growth behaviour in presence of pesticides
The main idea of our project was to use a biological organism as biosentinel which would warn us on the presence of contaminants on our tap water, analogically to the canaries in the coal mines, that would warn the miners in case of the release of toxic gases.
Our choice of organism was yeast (S. cerevisiae). However, sensitivity is one of the biggest issue of our project. As a consequence, we needed to find a way to make yeast more sensitive. Our idea was to look into the yeast stress response pathway . This pathway is controlled by the master regulators MSN2 and MSN4, as well as the YAP1 transcription factor. Our hypothesis was that pesticides would activate the stress response of our organisms and their survival mechanisms: deleting the master regulators would then weaken the strains and affect the growth of the cells. The next step was verifying our hypothesis.
Our first results show that in presence of a solution of pesticides dissolved in Methanol (10%) with a concentration of 1mg/l, the deletion strains reveal a delay in the initiation of the growth of the cells , as well as a change in the growth rate compared to the wild type strain in the same conditions.
The next question we had was whether or not the induced effect on the growth behavior was due to the pesticides alone, or if the Methanol in which the pesticides were dissolved was playing a role. By only looking at the growth curves, we did not succeed in addressing this matter, as the growth behavior of the deletion strains in presence of pesticides dissolved in Methanol and Methanol alone was similar.
Figure 1: Summary of MSN2 deletion strains results
This uncertainty led us to a more theoretical approach. We modeled the different curves by a logistic function and looked at the growth factor. For each condition (representing a strain - pesticide pair), we had 4 replicates, which allowed us to apply a Student’s t-test corrected by the Benjamini-Hochberg procedure on the samples, in order to see if we observed a significant difference in the mean growth factor between pesticides solutions and Methanol alone.
Our results show that the MSN2 deletion strain has a significantly different growth factor in presence of 2 pesticides, Diazinon and Metolachlor, in comparison with the solution of Methanol alone. To conclude, by deleting the gene coding for the MSN2 master regulator, yeast cells present a different growth behavior in the presence of pesticides. If those strains were put in presence of contaminated water samples, we would be able to detect this contamination by measuring their growth behavior and comparing it to a baseline.
Inducible expression of the fluorescent reporter by HSP12 strain
We integrated the mScarlet-I reporter protein under the control of the promoter of the stress response gene HSP12 in the genome of wildtype yeast. The successful integration was confirmed by sequencing of the target locus. The data we generated under the fluorescence microscope shows that the reporter protein is expressed.
Figure 2: In a plate reader assay we found that the fluorescence signal after reaching the plateau phase is higher if the cells are grown in the presence of the herbicide metolachlor. In spite of some variance in the experiment this is a promising result that we will follow-upon with more screening experiments.
Yeast viability in the capsule
A key point of our project is the capsule sitting inside the hardware box. Indeed, in our proposed implementation we plan to deliver these capsules already filled with our genetically modified yeast, which will have been freeze-dried prior to the delivery and mixed with SC medium in powder. For this, we had to prove that it was possible for dry yeast cells to be rehydrated and grow at a correct rate inside the capsules that we designed. More details on the capsule can be found here.
Commercial baking dry yeast was used along with SC medium inside a capsule. 4 samples were taken at the beginning, and again after 16 hours and the optical density was measured using a plate reader. The Table 1 below shows the OD600 values that were measured and the average.
Table 1: Measurements of OD600 at time t=0 and t=16h of yeast cells in our capsule.
A factor of 5.5 was calculated between these two values. The figures below show the capsule before and after the 16 hours and a visible difference can be seen in the transparence of the solution.
With this we have demonstrated that it was possible for dry yeast to grow inside the vials that we designed for Espress'EAU.
Figures 3 and 4: Growth of commercial dry yeast inside the capsule. On the left before, on the right after 16 hours.
Hardware implementation
After working on all the separated components, from the capsule to the detectors and the 3D printed parts, the whole device was assembled into a prototype. The last tests that we did were to prove that it was functional and able to measure both a fluorescent signal and the optical density during several hours.
For this, a growth curve experiment was conducted by the fully integrated hardware with the use of a supermarket yeast using SC Medium.
The experiment was conducted overnight and was ran for 16.9 hours with the data being stored and processed using Python code. The sensor gives us readings in terms of the light transmission which were converted into OD600 readings. The integration shows us that as time increases, the density of the cells increases which acts as a valid results for the proof of concept. The light transmission measurements and the conversion to optical density were verified by comparing the results to those from a spectrophotometer prior to this experiment. More details on the optical density sensor components and the corresponding results can be found here.
Some minor instabilities can be observed in the curve below. This can be due to many factors such as not full enclosure of the box which isolates the hardware, extra load on the circuit containing the circuit/Peltier or a biology artefact, which may have caused the instability of the environment. Nevertheless, even with these small defects a correct growth curve was recorded by our device, showing that the optical density detector works correctly.
Figure 5: Growth curve of yeast cells measured by our sensor
Finally, the fluorescence sensor had to be tested too in order to confirm that it is possible to use our Espress’EAU to measure the fluorescence emitted by the mScarlet-I protein expressed by the reporter yeast strains. As we were racing against the clock and we wanted to minimize the possible error sources, we decided to rely on an inducible strain for these tests. The fluorescence was tested for three different colors (red, green and blue) using a RGB light sensor and only a red fluorescent signal was expected. Different LED resistances were also used for these tests.
Figure 6: Fluorescence measurements of inducible yeast strains in three different wavelengths
These curves show that indeed, a significantly stronger fluorescent signal was recorded when detecting the red fluorescent emission compared to the other wavelengths. Water was used as a negative control and showed a low fluorescence intensity.
These results confirm that the fluorescence sensor that has been integrated in our device can be used to measure the emissions of the yeast cells contained in the capsule and let us predict that a similar result should be obtained with the reporter strains that we engineered.
Conclusion
We have shown that some of our engineered yeast strains have the intended effect: in the presence of pesticides, some deletion strains show a growth behavior that is distinguishable from the control and one reporter strain produces a higher fluorescence signal. Furthermore, we have shown that our device is able to measure both optical density and fluorescence. Given these results, we are confident that in the future, Espress’EAU will be capable of differentiating between a clean water sample and one that is polluted with pesticides, and possibly other contaminants.
Special thanks to our sponsors!
Follow us on our social media.
@EPFL iGEM 2020