Introduction
The goal of our project was to manufacture a working device that could reliably detect the presence of gamma-hydroxybutyric acid (GHB) in drinks. We developed a transcription factor-based capacitive biosensor that incorporated: DNA, transcription factor, electrodes, and hardware to monitor a change in GHB concentration in drinks. After finding a transcription factor for GHB, known as BlcR, we decided to use its dissociation property to create a biosensor capable of detecting the drug. We envisioned a sensor in which oligonucleotides are immobilized on a gold plate. The DNA binds with the BlcR, a dimer that tetramerizes when in contact with DNA. In the presence of GHB, the transcription factor dissociates from the DNA and binds GHB. This release generates a capacitance change, which can be translated into an electrical signal. This mechanism, shown in Figure 1, is the basis of our capacitive biosensor. In addition, we wanted to create a sensor as reliable as possible, for this reason we investigated the optimal threshold value.
Figure 1. Systematic overview of our GHB detecting sensor. A DNA sequence specific to the binding of the transcription factor BlcR is immobilized on an electrode that can measure capacitance differences. In the absence of GHB, BlcR is bound to the DNA resulting in a more dense environment between the electrodes. In the presence of GHB BlcR will dissociate from the DNA resulting in a change in the environment which can be converted to an electrical signal.
Effective dissociation
To develop a biosensor with these characteristics, we first had to make sure that the dissociation between the immobilized DNA and the transcription factor caused by the presence of GHB was occurring. Unfortunately, we couldn’t perform all the experiments with GHB since it is a controlled substance. Alternatively, we used an analog for GHB known as succinic semialdehyde (SSA) [TU Delft, see safety ]. We did this using Atomic Force Microscopy (AFM), which analyzes surfaces through a tip and a cantilever. We measured a gold plate with four different conditions: an empty gold plate (Figure 2a), then immobilized with 1 µM thiol-modified DNA (Figure 2b), then immobilized with 3 µM BlcR (Figure 2c) and finally 15 µM SSA added (Figure 2d) [TU Delft, see Notebook 4 , 09.09.2022].
Figure 2. AFM image of (a) an empty gold plate, (b) a gold plate immobilized with 1 µM thiol modified DNA, (c) a gold plate immobilized with 1 µM thiol modified DNA, with 3 µM BlcR and (d) a gold plate immobilized with 1 µM thiol modified DNA, with 3 µM BlcR, with 15 µM SSA. All images are scaled to a maximum value of 35 nm.
According to the results, BlcR increases the number of dots on the surface, indicating that the protein is bound to the DNA strand [Figure 2c]. Last but not least, the surface has significantly less dots when SSA is introduced, which makes us believe that dissociation took place [Figure 2d].
To obtain quantitative information from the images produced by the AFM, we used ImageJ to determine the amount of dots per image. The dots that fall beneath the applicable threshold of 18–255 have been counted. This has been done with smoothing, a practice that partially eliminates pictures' noise, see Table 1.
Table 1. The results of this action for the images produced with AFM, with smoothing. The values corresponding with the empty gold plate are marked under ‘empty’. The values corresponding with the gold plate immobilized with 1 µM thiol modified DNA are marked under ‘DNA’. The values corresponding with the gold plate immobilized with 1 µM thiol modified DNA, with 3 µM BlcR are marked under ‘BlcR’. The values corresponding with the gold plate immobilized with 1 µM thiol modified DNA, with 3 µM BlcR, with 15 µM SSA are marked under ‘SSA’.
|
Empty (dots) |
DNA (dots) |
BlcR (dots) |
SSA (dots) |
| With smoothing |
x |
9 |
366 |
197 |
From the data in Table 1 we can state that more dots are visible after BlcR has been immobilized on the electrode. Additionally, exposure to SSA reduces the number of dots by about 40%. The rise in number of dots compared to the values under "DNA" and the decrease in the number of dots compared to the values under "BlcR" may support the hypothesis that BlcR binds to immobilized DNA and that SSA unbinds BlcR.
Functioning capacitive biosensor
Once the dissociation was confirmed, we measured the capacitance generated by BlcR binding and unbinding to try and understand if the position of BlcR could generate a capacitance change large enough to be accurately measured [TU Delft, Notebook 4 , 23.08.2022].
Vector impedance
To exclude capacitance fluctuations brought on by circumstances other than BlcR binding, a two-electrode method has been developed to monitor the change in capacitance brought on by SSA addition and, consequently, the binding and unbinding of BlcR. The wild-type BlcR binding site-containing oligonucleotides have been immobilized on electrode A. Scrambled oligonucleotides (lacking the BlcR binding site) were immobilized on electrode B. See the protocol section for more information on DNA immobilization and electrode preparation. The electrodes have been examined using vector impedance measurements at 1 kHz after DNA immobilization. Figure 3 shows the average capacitance of electrodes A and B with DNA, BlcR, and SSA.
Figure 3. Visualization of the average capacitance during the vector impedance measurements. The capacitance of (a) electrode A (with BlcR binding site on DNA) and (b) electrode B (without BlcR binding site on DNA) has been measured without BlcR or SSA addition, labeled ‘DNA’, after incubation with BlcR, labeled ‘BlcR’, and after incubation with SSA, labeled ‘SSA’.
Figure 3 shows that when BlcR is added to electrode A, the capacitance reduces from an average of 1.195 pF to 0.803 pF, indicating that the BlcR is attached to DNA. However, when SSA is added, the capacitance increases to 0.899 pF, indicating that SSA binds to BlcR, which causes the dissociation of BlcR from the DNA. Since the capacitance does not equal the starting capacitance value, we can also assume that the SSA addition does not completely unbind BlcR. With electrode B, we can observe that adding BlcR or SSA has little effect on capacitance, causing a change of 2.34% and 5.25% respectively. The exact values can be found in Notebook 4 , 23.08.2022.
Universal Transducer Interface
We also utilized a the Universal Transducer Interface (UTI) in addition to the vector impedance measurements. Employing two reference capacitors, a UTI was utilized to measure the capacitance with greater accuracy. The system first uses capacitance-to-voltage converter to convert the capacitance to a voltage (CVC). A voltage-to-period converter is then used to translate this voltage signal into a particular period (VPC). Lastly, the frequency is adjusted to the required data acquisition frequency using a digital divider.
We conducted an experiment that was comparable to the one with vector impedance [TU Delft, Notebook 4 , 24.08.2022]. For each of the three conditions: electrode with DNA; electrode with DNA and BlcR, electrode with DNA, BlcR, and SSA, we averaged 50 capacitance measurements. Figure 4 present the findings.
Figure 4. Visualization of the average capacitance during the UTI measurements. The capacitance of (a) electrode A (with BlcR binding site on DNA) and (b) electrode B (without BlcR binding site on DNA) has been measured without BlcR or SSA addition, labeled ‘DNA’, after incubation with BlcR, labeled ‘BlcR’, and after incubation with SSA, labeled ‘SSA’.
The outcomes of the UTI measurements and the vector impedance measurements are comparable. When BlcR is added, electrode A's capacitance decreases from 0.7129 pF to 0.5602 pF; when SSA is added, it increases to 0.6421 pF. All values can be found in Notebook 4 , 24.08.2022. Once BlcR or SSA are added to electrode B, the capacitance does not change significantly. These results support the theory that when DNA with a BlcR binding site is immobilized, BlcR binding reduces capacitance and SSA addition increases capacitance. By comparing electrodes A and B, it is possible to design a system where the capacitance is altered when SSA is added while minimizing the impacts of other external factors (such as temperature or salt concentration variations).
These findings allowed us to establish the viability of developing a capacitive biosensor based on transcription factors. Our ability to submit a priority patent application for this technology was possible thanks to the novelty of our project.
Arduino measurements
The final method we used to measure a change in capacitance was Arduino. More information about our Arduino circuit can be found in the hardware page. We observed during the Arduino experiments performed throughout our project (see Notebook 4) that the addition of SSA resulted in a shift of about 6 relative values. The highest amount of noise was under 0.5 relative values. As a compromise, we selected a cutoff value for the light of 3 relative values, which results in a capacitance shift of 1.6*10-2 pF. This means that for capacitance changes higher than the mentioned value the LED light in our system will be turned on [see the hardware page for more information].
For this experiment, electrodes A and B have been prepared with 1 µM DNA and incubated with 4 µM BlcR for 1 hour [TU Delft, Notebook 4 , 27.09.2022]. After that, the electrodes were connected to the circuit and the relative output was measured. With a calibration line (presented in Figure 5), these values have been converted to pF units. This experiment has been performed in duplo. See Figure 6.
Figure 5. Validation experiments for the read-out circuit.
Figure 6. Results of the Arduino capacitance measurements. The calculated capacitance of electrode A and B when SSA was added to the solution with BlcR. After BlcR incubation for 1 hour, the output was baselined (set to zero).
GHB results
Since TU Delft lacks an authorization to work with GHB, the unbinding of BlcR from the immobilized DNA has been induced by SSA (GHB analog) until now. However, we had the opportunity to test GHB at the Trimbos Institute, which does have the permit. An electrode A and an electrodes B (1 µM DNA) were analyzed after incubation with 4 µM BlcR and after addition of GHB to a final concentration of 5 g/L (standard concentration of GHB in spiked drink [1] ) [TU Delft, Notebook 4 , 04.10.2022]. See Figure 7.
Figure 7. Visualization of the average capacitance during UTI measurements. (a) Electrode A with 1 µM immobilized Blc operator. (c) Electrode B with 1 µM immobilized scrambled DNA. Capacitance has been measured after incubation with 4 µM BlcR: ‘BlcR’ and after addition of GHB to an end concentration of 55 mM GHB: ‘GHB’ (Notebook 4, 04/10/2022).
Once again, the BlcR ligand (GHB) increases the capacitance of electrode A by around 0.025 pF, which is higher than our cutoff value of 0.016 pF. With electrode B, the capacitance does not change significantly.
Conclusion
Our project's objective was to create a functional device that could accurately detect gamma-hydroxybutyric acid (GHB) in beverages. To track changes in GHB content in beverages, we created a transcription factor-based capacitive biosensor that included DNA, transcription factor, electrodes, and hardware. We based our sensor on a technique that exploits a transcription factor's binding and unbinding to DNA to detect changes in capacitance. In order to avoid false positives we set a threshold value which was then confirmed by our experiments. We may say that this procedure occurs in practice because of the aforementioned experiments.