To develop our proof of concept we needed to accomplish four main goals:

1. Chemically modify the surface of capillary tubes
2. Immobilize DNA strands on the surface of capillary tubes
3. Verify toe-hold mediated strand displacement reactions (TMSDR) on glass surfaces
4. Modify the wettability of the inside of a capillary tube in order to observe a visible change in water rise

A detailed description of the methods can be found in the Protocols section of the wiki.

GLASS FUNCTIONALIZATION

In order to prepare a TMSDR in a capillary tube, the substrate strand needs to be immobilized on glass. Modifying the surface with a chemical group that can be used to covalently bind the DNA to the surface is integral. The chemical 3-Aminopropyltriethoxysilane (APTES) is often used to functionalize these surfaces with primary amines for crosslinking other reagents to glass [1]. To optimize the reaction, the choice of solvent and temperature were tested as they are of special relevance to the reaction kinetics and monolayer quality [1]. Different solvents interact differently with water. The nature of these interactions will dictate the yield and speed of the reaction as water can either inhibit or promote the reaction at different concentrations [1].

Success of the reaction can be assessed through ninhydrin assays, water contact angle (WCA) measurements, and fluorescence assays. Ninhydrin is a crystalline compound that reacts with the amine groups of APTES to form a product known as Ruhemann's purple (RP), which absorbs light at 570 nm [2]. WCAs serve as an indicator of the wettability of the glass surface, which is significantly altered due to APTES functionalization [1]. Fluorescein isothiocyanate (FITC) is a fluorophore that can be used to label primary amine, meaning that fluorescence would confirm the presence of these groups.

Figure 1. Methods employed for verification of amine functionalization on glass surfaces.

The reaction was tested with anhydrous toluene, ethanol, and a mixture of toluene and ethanol in a 9:1 ratio. Similarly, a second group of tests was done with the reaction at room temperature, 50°C or 70°C. All the reactions were done overnight. The products of the ninhydrin test showed significant absorption of light at 570nm indicating a successful reaction.

Figure 2. Ninhydrin tests absorbance values of glass slides functionalized with APTES in Ethanol (top), Toluene (bottom) and Ethanol+Toluene (middle).

Likewise, the contact angles for all the groups showed a significant difference from the negative control:

Figure 3. Water droplets on reacted (left) and unreacted (right) glass slides.

Figure 4. Mean measurements of water contact angle for glass slides reacted with APTES under different temperatures. All experimental groups had a significantly higher contact angle than the negative control (n=5, M=25.40°, SD=7.92). The 50°C (n=5, M=39.33°, SD=6.66°) and 70°C (n=5, M=36.90°, SD=3.13°) were both significantly lower than the room temperature (n=5, M=55.40°, SD=4.04°) group.

Figure 5. Mean measurements of water contact angle for glass slides reacted with APTES under different anhydrous solvents. All experimental groups were significantly higher from the unreacted glass slide group (n=5, M=25.4°, SD=7.93°). The ethanol (n=5, M=64.1°, SD=4.62°), toluene (n=5, M=59.0°, SD=8.72°), and 9:1 toluene/ethanol (n=5, M=61.3°, SD = 4.97°)

The measured WCA are consistent with what is expected from aminated surfaces based on current literature [1]. Although all the reactions were successful, the different solvent groups were not found to be significantly different from each other. Similarly, incubation at 50°C or 70°C did not produce any significant differences either. Surprisingly, incubation at room temperature seemed to produce significantly more functionalization than any other temperature. However, this is likely due to significant physioadsorption to the glass rather than covalent immobilization as reported by other studies [1].

Based on our results, and the scheme described by Sun et al. [3], we decided to modify the glass surfaces at 60°C for 1 hour using a 10% APTES solution in ethanol. The reaction was first tested inside scintillation vials, then glass slides, and finally capillary tubes to ensure that if there were any problems at any point, it was not due to the reaction itself. FITC fluorescence was tested using a UV-lamp (365 nm) in a dark room using PBS Buffer as a visualizing solvent. All the experimental groups exhibited visible fluorescence while the controls were mostly dark:

Figure 6. Fluorescence of glass vials, slides, and capillary tubes visualized under a UV-lamp. Substrates on the left were reacted with APTES (10%) for 1 hour at 60°C and then with FITC in DMF under the same conditions. Substrates on the right are unreacted. PBS buffer was used as the medium for the images.

By successfully functionalizing capillary tubes with primary amines, we checked off one of our main goals of the project and could move to the next steps of assembling our sensor.

DNA IMMOBILIZATION

The team decided to follow the same immobilization strategy as Sun et al. [3]. First, capillary tubes were reacted with APTES to functionalize the surface with primary amines. Next, the surface was reacted with Glutaraldehyde to get amine-reactive groups (aldehydes) on the inside of a capillary. Finally, DNA oligonucleotides modified with a primary amine group on the 3' end (purchased from IDT), were reacted onto the glass to finally immobilize the substrate strand on our capillary tubes.

These series of reactions were verified using DNA strands modified with fluorescent molecules which were then used to visualize the surface of a modified capillary tube under a confocal microscope. The data shows a statistically significant signal for the reacted capillary tube when compared to the control (n = 3, M = 1.31 A.U, SD = 0.22, p = 0.016 versus control, Student's t-test).

Figure 7. Confocal microscopy imaging and average count for 560 ± 94 nm fluorescence of capillary tubes functionalized (n=3, M=1.31 A.U, SD=0.22) and unreacted (n=3, M=0.24, SD=0.09) with DNA strands modified with Cy3. Capillary tubes are highlighted in the red box.

As part of our work, we also tested a modified version of a method described by Wickramathilaka & Tao to prepare oligonucleotides [4] functionalized on the 5' end with primary amines. Primary amines are one of if not the most useful group for chemical crosslinking of biological agents [1], which means functionalizing DNA strands with such a group can prove incredibly useful for any synthetic biology projects involving any modifications of oligonucleotides. To verify this reaction, FITC was reacted with the modified DNA mixture after which it was purified with an NEB Monarch® PCR & DNA Cleanup Kit. The purified DNA was then analyzed with a fluorometer, which showed a much higher signal when compared to controls.

Figure 8. Fluorescence spectra (λex = 495 nm) of samples with DNA strands functionalized with FITC, purified through a silica column (left); and samples containing all the components of the functionalization reaction except DNA, purified through a silica column (right).

Two graphs on the right represent a control, two on the right represent the purified DNA solutions.

We also tested if these modified DNA molecules could be immobilized onto glass surfaces for use in biosensors. For this, double stranded DNA was functionalized with primary amines, then reacted with glutaraldehyde-functionalized capillary tubes, and then reacted with FITC. When visualized under a UV Lamp, fluorescence was observed with the naked eye, indicating a successful reaction.

Figure 9. Fluorescence of glass capillary tube functionalized with FITC-modified DNA strands Under a UV Lamp.

We hope that with this procedure, future teams working on similar projects can easily immobilize oligonucleotides onto glass surfaces using minimal reagents and without the need to purchase costly custom oligos.

TOEHOLD MEDIATED STRAND DISPLACEMENTS

TMSDs are the sensing mechanism of our proposed device, so it is crucial to ensure that we can successfully run this process on glass surfaces. For this purpose, we needed to integrate everything that we have accomplished up to this point into a functional biosensor. The glutaraldehyde-functionalized glass slides were used to immobilize the substrate strands and set the platform for the TMSDs to occur. The substrate strand was modified in the 5' end with the Cy3 fluorophore and in the 3' end with a primary amine. The 3' end was chosen as the attachment point as previous studies show that the TMSD is much more efficient when the toehold is in the 5' end of the substrate strand [4]. The incumbent strand was modified in the 3' end with Cy5, a dye capable of acting as a Fluorescence Resonance Energy Transfer (FRET) acceptor for Cy3. By acting as a FRET acceptor, we can monitor when the substrate and incumbent strands are hybridized together. The target strand was modified in the 3' end with 6-FAM, a fluorophore similar to Cy3.

Figure 10. Reaction scheme for toehold-mediated strand displacements on glass surfaces.

Three slides were prepared and modified up to different steps of the reaction scheme. The first slide was functionalized with the substrate and then imaged with the microscope. The second slide was hybridized with the incumbent strand and then imaged. The final slide was treated with 100 nM of the target strand, and then imaged.

Figure 11. Confocal microscope imaging (λex = 470/635 nm) of glass slides functionalized with a DNA Substrate modified with Cy3. The left image corresponds to a channel measuring fluorescence of light >635 nm (Cy5 emission) and the right image corresponds to a channel measuring fluorescence of light in the 560±94 nm range (Cy3/6-FAM emission).

Figure 12. Confocal microscope imaging (λex = 470/635 nm) of glass slides functionalized with a DNA Substrate modified with Cy3 and hybridized with a DNA incumbent strand modified with Cy5. The left image corresponds to a channel measuring fluorescence of light >635 nm (Cy5 emission) and the right image corresponds to a channel measuring fluorescence of light in the 560±94 nm range (Cy3/6-FAM emission).

Figure 13. Confocal microscope imaging (λex = 470/635 nm) of glass slides functionalized with a DNA Substrate modified with Cy3 and hybridized with a DNA incumbent strand modified with Cy5. The left image corresponds to a channel measuring fluorescence of light >635 nm (Cy5 emission) and the right image corresponds to a channel measuring fluorescence of light in the 560±94 nm range (Cy3/6-FAM emission). 6-FAM not included in the diagram for simplicity.

These results showed that the target sequence is capable of displacing the incumbent strand from the complex at a concentration of 100 nM, a process that can be observed within minutes.

MODIFYING WETTABILITY

Our entire proposed project depends on producing a big enough change in wettability to observe a signal easy to observe. To test how much we can modify the wettability of a capillary tube -and by proxy the water rise due to capillary action- a chemical called Octadecyltrichlorosilane (OTS) was reacted with the surface of different tubes and then the heights of the water from the surface were measured. OTS is highly reactive towards glass surfaces and produces a highly hydrophobic layer as a result [5]. This was confirmed with our experiments which showed that when dipped in water, no liquid would even go into the capillary.

Figure 14. Glass capillary tubes dipped in water. The left image shows a capillary functionalized with OTS, the right image shows an unreacted capillary. Water rise was only observed in the unreacted tube, while water was suppressed for the functionalized tube.

CONCLUSIONS

Our team achieved strong results developing a proof-of-concept biosensor. We functionalized glass surfaces with DNA oligos and altered capillary tube wettability enough to observe a visible signal. This allowed us to assemble a working biosensor and establish methods for advancing the project. Next, we aim to generate a wettability change as a signal for detecting specific DNA sequences. To do this, we'll modify oligos with hydrophobic groups now feasible thanks to our successful DNA modification tests. We'll replace fluorophores with hydrophobic tags in the TMSD system, test sensitivity, and optimize detection limits for point-of-care use. These results mark a promising beginning for Rumino's journey ahead.

REFERENCES

[1] Sypabekova, M., Hagemann, A., Rho, D., & Kim, S. (2022). Review: 3-Aminopropyltriethoxysilane (APTES) Deposition Methods on Oxide Surfaces in Solution and Vapor Phases for Biosensing Applications. Biosensors, 13(1), 36. DOI: 10.3390/bios13010036

[2] Stauß, A. C., Fuchs, C., Jansen, P., Repert, S., Alcock, K., Ludewig, S., & Rozhon, W. (2024). The Ninhydrin Reaction Revisited: Optimisation and Application for Quantification of Free Amino Acids. Molecules (Basel, Switzerland), 29(14), 3262. DOI: 10.3390/molecules29143262

[3] Sun, Y., Liu, K., Zhang, H., Zhao, Y., Wen, J., Zhao, M., Li, X., & Li, Z. (2025). A tube-based biosensor for DNA and RNA detection. Science Advances, 11(18), eadu2271. DOI: 10.1126/sciadv.adu2271

[4] Wickramathilaka, M. P., & Tao, B. Y. (2019). Characterization of covalent crosslinking strategies for synthesizing DNA-based bioconjugates. Journal of Biological Engineering, 13, 63. DOI: 10.1186/s13036-019-0191-2

[5] McGovern, M. E., Kallury, K. M. R., & Thompson, M. (1994). Role of Solvent on the Silanization of Glass with Octadecyltrichlorosilane. Langmuir, 10(10), 3607-3614. DOI: 10.1021/la00022a038