Liquid-to-gel transition for visual and tactile detection of biological analytes

Tatiana A. Fedotova a and Dmitry M. Kolpashchikov *abc
aChemistry Department, University of Central Florida, Orlando, 32816, Florida, USA. E-mail:
bBurnett School of Biomedical Sciences, University of Central Florida, Orlando, 32816, Florida, USA
cITMO University, Laboratory of Solution Chemistry of Advanced Materials and Technologies, Lomonosova St. 9, 191002, St. Petersburg, Russian Federation

Received 7th September 2017 , Accepted 23rd October 2017

First published on 23rd October 2017

So far all visual and instrument-free methods have been based on a color change. However, colorimetric assays cannot be used by blind or color-blind people. Here we introduce a liquid-to-gel transition as a general output platform. The signal output (a piece of gel) can be unambiguously distinguished from liquid both visually and by touch. This approach promises to contribute to the development of an accessible environment for visually impaired persons.

Among the variety of analytic methods, those that can produce a direct visual output are particularly attractive due to the possibility of instrument-free signal readout.1 Indeed, analytical formats such as pH test strips, the home pregnancy test, enzyme-linked immunosorbent assay (ELISA) or gold-nanoparticle-based DNA analysis2 can be said to be the best in terms of the easiness of output readout. Here we challenge this statement by introducing an alternative signal readout for molecular sensors, which can be detected not only visually but also by touch. The tactile signal readout can be used by blind and color-blind people or under conditions when visualization is not possible (e.g. in the dark).

This method is based on the radical-mediated enzymatic polymerization of acrylamide into polyacrylamide by a horseradish peroxidase (HRP)/acetylacetone/H2O2 ternary system.3 In this system acetylacetone is oxidized by HRP to the acetylacetone-radical, which results in the radical-initiated polymerization of acrylamide accompanied by gel-to-liquid transition. However, HRP itself cannot be transformed easily into a sensor capable of detecting biological analytes. On the other hand, a number of DNA and RNA sequences were reported to have HRP-like activity in the presence of hemin.4 The peroxidase-like DNAzymes (PxD) contain guanine-rich DNA sequences that fold into a stable G-quadruplex structure, which can bind hemin and catalyze H2O2-dependent oxidation of colorless substances and convert them into colored products. Furthermore, a great variety of colorimetric biosensors have been developed based on PxD.5 Here, we aimed at achieving an analyte-dependent activation of PxD followed by the polymerization of acrylamide solution rather than a color change. If polymerization is completed, the gel fragments would be observed at the tips of the inversed tubes, or by touch if the contents of the tubes are released onto filter paper (Fig. 1). In this proof of concept study, we report on the detection of biologically important analytes: ATP and two DNA sequences by a liquid-to-gel transition method.

image file: c7cc07035g-f1.tif
Fig. 1 Polymerization-based visual and tactile detection of ATP. (A) Sensor design: DNA strands PxD1 and PxD2 bind two ATP molecules,6 which stabilize a G-quadruplex structure. The G-quadruplex binds hemin (brown oval) and catalyzes the H2O2-dependent oxidation of acetylacetone to the acetylacetone radical, which then triggers the polymerization of acrylamide and bisacrylamide. Dotted lines represent triethylene glycol linkers. (B) The visual signal output can be detected by turning the reaction test-tube upside down. All samples contained reaction buffer: 50 mM HEPES pH 6.6, 50 mM MgCl2, 20 mM K+, 120 mM NaCl, 1% DMSO v/v, 0.03% w/v Triton × 100, 38% w/v acrylamide, 2% w/v bisacrylamide, 127 mM acetylacetone, 5 μM hemin and 5.2 mM H2O2. Polymerization was observed only when either 5 or 10 mM of ATP were present (samples 3 and 4, respectively). No gel fragments are observed at the tip of the tube in the absence of ATP (sample 2). Sample 1 (negative control) contained only the buffer components; sample 5 (positive control) contained full PxD 5’-GGG TAG GGC GGG TT GGG. lower panel: tactile readout is possible if the samples are placed on a piece of filter paper. Liquid samples 1 and 2 are absorbed by the paper, while gel shapes from samples 3, 4 and 5 can be sensed by touch. All samples were cyan colored due to the addition of 0.1% w/v xylene cyanol for easy visualization.

We first turned our attention to the detection of ATP, an important biological analyte. The sequence for an ATP aptamer was isolated by Huizenga and Szostak.6a The aptamer is known to bind two ATP molecules and was used in a number of assays for ATP detection.6 In our design, the ATP aptamer sequence was split into two fragments, PxD1 and PxD2 (Fig. 1A). Each half was attached to a part of the G-rich PxD sequence. Binding of ATP to the aptamer stabilized the G-quadruplex structure, which catalyzed the polymerization reaction in the presence of hemin. To prevent nonspecific association of the sensor strands in the absence of the analyte, the PxD2 strand had additional nucleotides, allowing the formation of a stable stem–loop structure within one of the sensor strands (Fig. S1, ESI). Five mM ATP was detected in 1 h of incubation at room temperature (Fig. 1B, sample 3). This concentration is within the physiological range of the ATP level.7 The addition of 10 mM ATP resulted in polymerization within ∼30 min (Fig. 1B, sample 4). No polymerization was observed in the presence of just two sensor strands (Fig. 1B, sample 2). Gel fragments were characterized under a microscope (Fig. S2, ESI).

The dependence of polymerization rates on the analyte concentration promised to become the basis for quantitative detection, in which the analyte concentration is determined based on the time required for polymerization. We performed a series of experiments aiming at determining a correlation between the ATP concentration and polymerization time. However, no clear dependence was determined in this study. Higher concentrations of the analyte triggered the polymerization of acrylamide at various times within one hour and varied from experiment to experiment. This can be explained by the presence of various amounts of residual oxygen in the samples, which is known to inhibit the polymerization reaction.8 This observation correlates with an irreproducible inhibition period for the HRP-mediated polymerization of acrylamide due to the presence of residual oxygen reported by Lalot et al.8 Due to the experimental setup, the presence of residual oxygen is inevitable and varies in each experiment from tube to tube, thus making it practically difficult to correlate the concentration of the analyte and polymerization time. Based on this observation, we would primarily recommend the presented method for qualitative analyte detection, when the presence or absence of an analyte (e.g. pathogen) in a sample needs to be determined. However, we do not exclude the fact that an improved oxygen-free setup will eventually enable applications of this method for quantitative analysis.

To demonstrate a general applicability of the approach, we used a similar strategy for a different biologically significant analyte, DNA. As a targeted sequence, we chose DNA fragments of AMEL_X and AMEL_Y genes, which encode for amelogenin – a protein involved in the development of tooth enamel. Different amelogenin-encoding sequences are located on the X and Y chromosomes, which is used for sex determination in forensic and anthropological applications.9 The split DNA sensor was designed as reported earlier for colorimetric DNA detection (Fig. 2A).10 Two strands, PxD3 and PxD4, hybridized to the abutting positions of the AMEL_X analyte and formed a PxD G-quadruplex structure, which triggered the polymerization reaction. In the absence of the analyte, no polymerization occurred (Fig. 2B, sample 2), whereas in the presence of 10 μM AMEL_X analyte liquid-to-gel transition was observed within 1 h (Fig. 2B, sample 3). Importantly, no polymerization took place in the presence of a single-base mismatched analyte (Fig. 2B, sample 4), which strongly suggests that this method can be used for the detection of point mutations and single nucleotide polymorphisms in DNA.

image file: c7cc07035g-f2.tif
Fig. 2 Polymerization-based visual and tactile detection of the AMEL_X DNA analyte. (A) Sensor design: two DNA strands PxD3 and PxD4 hybridize to a DNA analyte (AMEL_X) and form a G-quadruplex structure, which binds hemin and catalyzes the polymerization of acrylamide solution (see Fig. 1 legend and ESI for experimental details). Dotted lines represent triethylene glycol linkers. (B) Visual signal output. Polyacrylamide gel is observed in the presence of fully matched AMEL_X (3), but not in its absence (2) or in the presence of a mismatched analyte (4). Sample 5 did not contain DNA strands (negative control); sample 1 contained full PxD (positive control). Low panel: the gel fragments can be detected by touch after transfer onto filter paper.

Even though conventional hybridization probes can detect nucleic acid down to ∼1–100 nM,11 modern amplification technologies (e.g. LAMP, NASBA) produce highly concentrated nucleic acid amplicons. Furthermore, complete elimination of the contact between the assay and O2 can potentially further reduce the LOD of this assay. Therefore, the utility of the reported method in clinical DNA analysis has to be further validated.

Some analytical applications could benefit from the presence of an analyte-maintained liquid phase, while polymerization into gel occurred in the absence of the analyte. Therefore, we designed a sensor that detects a DNA analyte in the inversed (‘NOT’ logic) format, e.g. the presence of an analyte inhibits gel formation. We used the AMEL_Y DNA analyte in this design to demonstrate more general applicability of the assay for DNA detection. In this design, the catalytic DNA sequence was again split into two parts: PxD5 and PxD6 strands as shown in Fig. 3A. The two sensor strands associated and formed a G-quadruplex, which triggered gel formation in the absence of the analyte (Fig. 3B, sample 2). In contrast, the addition of the analyte disrupted G-quadruplex formation due to the preferential binding of the PxD5 strand to the AMEL_Y analyte. Consistent with the design, the polymerization of acrylamide was observed only in the presence of both sensor strands, but not the analyte (Fig. 3B).

image file: c7cc07035g-f3.tif
Fig. 3 Detection of the AMEL_Y DNA analyte in a NOT logic mode. (A) The predicted secondary structure of the AMEL_Y sensor in the absence of the AMEL_Y analyte. Nucleotides in green indicate a fragment complementary to the AMEL_Y analyte. The sensor acts in a ‘NOT’ format: it triggers polymerization in the absence of the AMEL-Y analyte. (B) Polymerization was observed in the absence of the AMEL_Y analyte (sample 2), but not in its presence (sample 3). Sample 1 contained full PxD (positive control); sample 4 contained all but DNA components (negative control). Lower panel: samples after release onto filter paper; only samples 1 and 2 can be detected by touch.

Methods for the user-friendly detection of biological analytes are under continuous development. For example, Macdonald and colleagues created a 7-segment display to vividly display an output of a molecular automaton that analyzed the DNA sequence of Ebola and Marburg viruses.12 Furthermore, Gormley et al. developed an analytical output to form polymers that can associate with multiple gold nanoparticles and form aggregates, resulting in a visible color change.13 Ikeda et al. reported peptide-based hydrogels that have the ability to encapsulate enzymes and detect small molecules including glucose, lactose and sarcosine.14 However, these and other approaches15,16 all relied only on the visual monitoring of the output signals. On the other hand, a great variety of electrosensors including smartphone-based, RGB-Depth, electromagnetic and others, have been developed for visually impaired people.17

Here we report the first method that can be used for the tactile detection of biological analytes and therefore can be used by blind and color blind persons without the need for any electronic device. Using this method, we were able to unambiguously detect ATP and two different DNA sequences (AMEL_X and AMEL_Y). Importantly, the DNA sequences can be detected with high selectivity even at room temperature, which is important in practice for the detection of drug-resistant bacteria and single nucleotide polymorphisms in human genomes. Importantly, PxD-based sensors are known for the detection of a broad variety of analytes including metal ions, small molecules and proteins.5 This method, therefore, has potentially a broad spectrum of bioanalytical applications.

The disadvantage of the reported assay is its sensitivity to atmospheric oxygen, which inhibits polymerization and complicates quantitative analysis. We do not exclude that future converting the technology into an oxygen-free format may solve this problem. The toxicity problem of unpolymerized acrylamide can be easily solved by wearing gloves or by covering the gel pieces by a plastic film. Alternatively, acrylamide can be replaced with less toxic monomers, e.g. hydroxyethyl methacrylate (HEA).18 Analysis of RNA or DNA sequences is not compatible with whole blood samples and, as for other nucleic acid detection technologies, requires isolation of DNA prior to analysis. However, we do not exclude that the method might be adoptable for some analytes (e.g. K+, drugs etc.) in whole blood, which would represent another major advance of the liquid-to-gel transition approach. We also believe that the new output signal reported in this study has great potential beyond a tactile-sensing platform. For example, liquid-to-gel transition can be used for analyte-dependent mechanical blocking of tubes or channels in (micro)fluidic devices thus regulating the liquid flow in an analyte-dependent manner.

We demonstrated an assay for analyte-dependent liquid-to-gel transition that can be sensed not only visually but also by touch. We hope that this method will make home testing systems accessible for blind and color-blind persons, as well as inspire future developments of chemical products for physically-challenged people.

Funding from NSF CCF 1423219 and NIH 1R15AI103880-01A1 is greatly appreciated. D. M. K. was partially supported by the ITMO University Fellowship and Professorship Program.

Conflicts of interest

There are no conflicts to declare.

Notes and references

  1. (a) M. J. Kangas, R. M. Burks, J. Atwater, R. M. Lukowicz, P. Williams and A. E. Holmes, Crit. Rev. Anal. Chem., 2017, 47, 138–153 CrossRef CAS PubMed; (b) Y. Zhang, I. D. McKelvie, R. W. Cattrall and S. D. Kolev, Talanta, 2016, 152, 410–422 CrossRef CAS PubMed; (c) G. Chen, Z. Guo, G. Zeng and L. Tang, Analyst, 2015, 140, 5400–5443 RSC; (d) A. Chen and S. Yang, Biosens. Bioelectron., 2015, 71, 230–242 CrossRef CAS PubMed; (e) M. S. Verma, J. L. Rogowski, L. Jones and F. X. Gu, Biotechnol. Adv., 2015, 33, 666–680 CrossRef CAS PubMed.
  2. (a) S. H. Radwan and H. M. Azzazy, Expert Rev. Mol. Diagn., 2009, 9, 511–524 CrossRef CAS PubMed; (b) H. I. Peng and B. L. Miller, Analyst, 2011, 136, 436–447 RSC; (c) Y. Zhou and J. Yoon, Chem. Soc. Rev., 2012, 41, 52–67 RSC; (d) B. T. Roembke, S. Nakayama and H. O. Sintim, Methods, 2013, 64, 185–198 CrossRef CAS PubMed; (e) F. Wang, L. Wang, X. Chen and J. Yoon, Chem. Soc. Rev., 2014, 43, 4312–4324 RSC; (f) R. J. Dinis-Oliveira, Bioanalysis, 2014, 6, 2877–2896 CrossRef CAS PubMed; (g) T. Pradhan, H. S. Jung, J. H. Jang, T. W. Kim, C. Kang and J. S. Kim, Chem. Soc. Rev., 2014, 43, 4684–4713 RSC; (h) C. Feng, S. Dai and L. Wang, Biosens. Bioelectron., 2014, 59, 64–74 CrossRef CAS PubMed; (i) M. J. Marín, C. L. Schofield, R. A. Field and D. A. Russell, Analyst, 2015, 140, 59–70 RSC.
  3. A. Durand, T. Lalot, M. Brigodiot and E. Marechal, Polymer, 2000, 41, 8183–8192 CrossRef CAS.
  4. P. Travascio, D. Sen and A. Bennet, Can. J. Chem., 2006, 84, 613–619 CrossRef CAS.
  5. (a) X. Yang, T. Li, B. Li and E. Wang, Analyst, 2010, 135, 71–75 RSC; (b) J. Kosman and B. Juskowiak, Anal. Chim. Acta, 2011, 707, 7–17 CrossRef CAS PubMed; (c) L. Neo, K. Kamaladasan and M. Uttamchandani, Curr. Pharm. Des., 2012, 18, 2048–2057 CrossRef PubMed; (d) B. T. Roembke, S. Nakayama and H. O. Sintim, Methods, 2013, 64, 185–198 CrossRef CAS PubMed; (e) D. Chang, S. Zakaria, M. Deng, N. Allen, K. Tram and Y. Li, Sensors, 2016, 16, pii: E2061 CrossRef PubMed; (f) D. L. Ma, C. Wu, Z. Z. Dong, W. S. Tam, S. W. Wong, C. Yang, G. D. Li and C. H. Leung, ChemPlusChem, 2017, 82, 8–17 CrossRef CAS; (g) K. Shahsavar, M. Hosseini, E. Shokri, M. R. Ganjali and H. Ju, Anal. Methods, 2017, 9, 4726–4731 RSC.
  6. (a) D. E. Huizenga and J. W. Szostak, Biochemistry, 1995, 34, 656–665 CrossRef CAS PubMed; (b) G. Zhu, M. Ye, M. J. Donovan, E. Song, Z. Zhao and W. Tan, Chem. Commun., 2012, 48, 10472–10480 RSC; (c) S. Ng, H. S. Lim, Q. Ma and Z. Gao, Theranostics, 2016, 6, 1683–1702 CrossRef CAS PubMed; (d) X. Zheng, R. Peng, X. Jiang, Y. Wang, S. Xu, G. Ke, T. Fu, Q. Liu, S. Y. Huan and X. Zhang, Anal. Chem., 2017, 89, 10941–10947 CrossRef CAS PubMed; (e) Y. Li, L. Sun and Q. Zhao, Talanta, 2017, 174, 7–13 CrossRef CAS PubMed.
  7. E. K. Ainscow, S. Mirshamsi, T. Tang, M. L. Ashford and G. A. Rutter, J. Physiol., 2002, 544, 429–445 CrossRef CAS.
  8. O. Emery, T. Lalot, M. Brigodiot and E. Maréchal, J. Polym. Sci., Part A: Polym. Chem., 1997, 35, 3331–3333 CrossRef CAS.
  9. (a) S. Q. Yan, Y. M. Li, C. Y. Bai, P. C. Guo, S. Si, J. H. Sun and Z. H. Zhao, Genet. Mol. Res., 2015, 14, 16241–16246 CrossRef CAS PubMed; (b) B. A. Álvarez-Sandoval, L. R. Manzanilla and R. Montiel, PLoS One, 2014, 9, e104629 Search PubMed; (c) H. Nogami, H. Tsutsumi, T. Komuro and R. Mukoyama, Forensic Sci. Int.: Genet., 2008, 2, 349–353 CrossRef PubMed.
  10. D. M. Kolpashchikov, J. Am. Chem. Soc., 2008, 130, 2934–2935 CrossRef CAS PubMed.
  11. (a) D. M. Kolpashchikov, Scientifica, 2012, 2012, 928783 CrossRef PubMed; (b) D. M. Kolpashchikov, Chem. Rev., 2010, 110, 4709–4723 CrossRef CAS PubMed; (c) Y. V. Gerasimova and D. M. Kolpashchikov, Chem. Soc. Rev., 2014, 43, 6405–6438 RSC.
  12. J. E. Poje, T. Kastratovic, A. R. Macdonald, A. C. Guillermo, S. E. Troetti, O. J. Jabado, M. L. Fanning, D. Stefanovic and J. Macdonald, Angew. Chem., Int. Ed. Engl., 2014, 53, 9222–9225 CrossRef CAS PubMed.
  13. A. J. Gormley, R. Chapman and M. M. Stevens, Nano Lett., 2014, 14, 6368–6373 CrossRef CAS PubMed.
  14. M. Ikeda, T. Tanida, T. Yoshii, K. Kurotani, S. Onogi, K. Urayama and I. Hamachi, Nat. Chem., 2014, 6, 511–518 CrossRef CAS PubMed.
  15. E. J. Wee, T. Ha Ngo and M. Trau, Sci. Rep., 2015, 5, 15028 CrossRef CAS PubMed.
  16. G. Sicilia, C. Grainger-Boultby, N. Francini, J. P. Magnusson, A. O. Saeed, F. Fernández-Trillo, S. G. Spain and C. Alexander, Biomater. Sci., 2014, 2, 203–211 RSC.
  17. (a) M. J. Proulx, J. Gwinnutt, S. Dell'Erba, S. Levy-Tzedek, A. A. de Sousa and D. J. Brown, Restor. Neurol. Neurosci., 2015, 34, 29–44 Search PubMed; (b) K. Yang, K. Wang, R. Cheng, W. Hu, X. Huang and J. Bai, Sensors, 2017, 17, pii: E1890 CrossRef PubMed; (c) E. Ko and E. Y. Kim, Sensors, 2017, 17(8), pii: E1882 CrossRef PubMed; (d) P. H. Cheng, Assist. Technol., 2016, 28, 127–136 CrossRef PubMed.
  18. (a) A. Durand, T. Lalot, M. Brigodiot and E. Marechal, Polymer, 2001, 42, 5515–5521 CrossRef CAS; (b) R. A. Derango, L. Chiang, R. Dowbenko and J. G. Lasch, Biotechniques, 1992, 6, 523–526 CAS; (c) E. Morisbak, V. Ansteinsson and J. T. Samuelsen, Eur. J. Oral Sci., 2015, 123, 282–287 CrossRef CAS PubMed.


Electronic supplementary information (ESI) available: Detailed experimental procedure and the structure of PxD oligonucleotides. See DOI: 10.1039/c7cc07035g

This journal is © The Royal Society of Chemistry 2017