DTU-Biobuilders

Results

The aim of EndoSense was to develop and test a cell-free EDC biosensor using human nuclear receptors. This was achieved with the androgen receptor using multiple different DNA templates. The tests were done using the natural hormones of the receptors. Additionally, we aimed to test an in-cell biosensor for proof of concept. This was successfully tested using the natural hormones of the receptors and water samples from various sources.

Table of Contents

In-cell assay: Test of EDCs in water samples


The initial test of the in-cell biosensor consisted of creating dose response curves using the natural hormones for the receptors. The LacZ values were converted to a percentage of the maximum value for easy comparison between the curves. Unfortunately, we were not able to obtain progesterone to test the PR, and for MR we saw very high standard deviations and no significant response (data not shown). The four functional dose-response curves are shown in Figure 1 below.

Dose response curves
Figure 1: Dose-response curves for the in-cell assay. Percent maximal response is shown for easy comparison. The concentration of the added hormone is shown (not the concentration in the reaction). All measurements were done in triplicate and standard deviations are represented by error bars. The sample positive controls are shown from left to right; ERb, ERa, AR, GR.

Finally, the system was tested on four water samples. We tested two different types of bottled water, tap water, and a lake sample from a local lake. As a positive control, the natural hormones were added to the tap water sample to show that we can replicate the data from the dose-response curves. These positive controls are shown in Figure 1 and fit the dose-response curves. All data from the test of the water samples is shown in Table 1 below:

Table 1: Overview of percent maximal response for the in-cell assay on water samples. All measurements were done in triplicates and standard deviations are represented by error bars. Colors correspond to the receptors shown in Figure 1.
Bottled water #1 Bottled water #2 Tap water Lake water
ERa 1.3% ± 5.9% 5.3% ± 3.7% 5.3% ± 3.6% -0.8% ± 3.0%
ERb 2.4% ± 16.6% 1.4% ± 6.0% 5.3% ± 4.9% -1.8% ± 1.8%
GR -13% ± 50.9% 9.4% ± 11.3% 10.5% ± 10.8% -2.2% ± 31.5%
AR 15.7% ± 53.4% 28.9% ± 24.7% 28.3% ± 24.0% 33.7% ± 53.0%

None of the measurements displayed in Table 1 are significantly different from the lowest value on the corresponding dose-response curve in Figure 1. Therefore we can conclude that none of these water samples contain EDCs binding the tested receptors above our measurement threshold.

Cell-free Biosensor: EndoSense Proof of Concept on EDCs detection

During the development of the biosensor, we started by optimizing the DNA quantity, buffer conditions, and detection methods. Then we proceeded to test the ligand's influence and inhibitory effects. We observed that the ligand affected the fluorescence output even in the absence of the receptor. We accounted for this influence when testing the complete biosensor system with the receptor included, by relatively increasing the signal when the ligand was added.

We demonstrated the capability of detection of the EndoSense biosensor as proof of concept by using the androgen receptor, since it was the most recommended by our interactions in the human practices and our own custom DNA template: T7-HREmin-sB-T. We used the hormone, testosterone, a natural ligand of the receptors to prove the functioning.

The detection of the EDC is recognizable by the difference in the adjusted signal between the presence and absence of the ligand (Figure 2). The output signal is expressed as a standardized value as we included in each measurement a serial dilution of Fluorescein Sodium Salt (FSS), from the distribution kit, and normalized the values as Micromolar Equivalent Fluorescein (MEF).

Figure 2: Statistical significance of ligand detection
Figure 2: Difference in signal between presence and absence of the testosterone hormone (indicated as H, at 4 µM) with the androgen receptor (R), and the DNA template (at a concentration of 10 nM). For each time point, a statistical t-test was performed between the two shown measurements. The p-value for each test is shown as stars symbols above the top point.

As we saw an increase in the signal, opposite to the decrease we were expecting instead, we reconsidered the working mechanism of the biosensor. As our data would suggest, the EndoSense biosensor seems to behave in an OFF/ON way, meaning that without the EDC, there is a low output. On the other hand, the presence of the ligand will result in an increase of the signal. This working mechanism is shown in the Figure 3 below.

Figure 3: Statistical significance of ligand detection
Figure 3: OFF/ON behavior of the EndoSense biosensor. When the water sample does not contain EDCs, the fluorescence signal is inhibited. When EDCs are present, this will result in the transcription of the broccoli aptamer and ultimately a fluorescence signal.

From this analysis, we show how EndoSense can detect an EDC in the form of testosterone, with a high statistical significance. This proof of concept shows great improvement from the current detecting methods, as the EndoSense biosensor can reduce EDC testing time from days to hours.

From our modeling analysis, we predicted the biosensor's Limit of Detection (LOD). We found an upper LOD of 8.39 µM and a lower LOD of 0.1 µM. While the current detection limit may be higher than ideal for real-world sample analysis, it has not yet been fully tested. Consequently, there is potential for improvement as we refine the system, since we have only done a few tests.

Back to top