Understanding signal and background in a ... - ACS Publications

Subscriber access provided by READING UNIV

Article

Understanding signal and background in a thermally resolved, single-branched DNA assay using square wave voltammetry Subramaniam Somasundaram, Mark D. Holtan, and Christopher J Easley Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b00036 • Publication Date (Web): 31 Jan 2018 Downloaded from http://pubs.acs.org on February 1, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Thermally controlled electrochemical cell allows precise optimization of a single-branched DNA assay using square-wave voltammetry. 30x11mm (300 x 300 DPI)

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Understanding signal and background in a thermally resolved, single-branched DNA assay using square wave voltammetry Subramaniam Somasundaram,a Mark D. Holtan,a and Christopher J. Easley*a * Corresponding Author a Department of Chemistry and Biochemistry, Auburn University, Auburn, USA E-Mail: chris.easley@auburn. Supporting Information Included ABSTRACT: Electrochemical bioanalytical sensors with oligonucleotide transducer molecules have been recently extended for quantifying a wide range of biomolecules, from small drugs to large proteins. Short DNA or RNA strands have gained attention recently due to the existence of circulating oligonucleotides in human blood, yet challenges remain for adequately sensing these targets at electrode surfaces. In this work, we have developed a quantitative electrochemical method which uses target-induced proximity of a single-branched DNA structure to drive hybridization at an electrode surface, with readout by square-wave voltammetry (SWV). Using custom instrumentation, we first show that precise control of temperature can provide both electrochemical signal amplification and background signal depreciation in SWV readout of small oligonucleotides. Next, we thoroughly compared 25 different combinations of binding energies by their signal-to-background ratios and differences. These data served as a guide to select the optimal parameters of binding energy, SWV frequency, and assay temperature. Finally, the influence of experimental workflow on the sensitivity and limit of detection (LOD) of the sensor is demonstrated. This study highlights the importance of precisely controlling temperature and SWV frequency in DNA-driven assays on electrode surfaces while also presenting a novel instrumental design for fine tuning of such systems.

In the diagnosis and treatment of genetic diseases, detection of oligonucleotides with high sensitivity and sequence specificity is of central importance1. Conventionally this has been accomplished through optical measurements during melting curve analysis of a polymerase-chain reaction (PCR) amplified oligonucleotide, where single-base mismatches can be resolved2. While a number of other specialized techniques exist, electrochemical measurement is an appealing alternative due to its small instrument size, low cost, and ease of handling1,2. In one common format for electrochemical oligonucleotide detection, a DNA/RNA capture probe which has a redox moiety like ferrocene or methylene blue (MB) is immobilized onto an electrode surface. Introducing the target, which has complimentary sequence to the capture DNA, results in a conformational change that affects the distance between the redox moiety and the surface and effectively alters the electron transfer rate to the electrode3–9. In these protocols, the percentage change in the signal is proportional to the target concentration. Similarly, there are methods in which the target brings a redox moiety to the surface from bulk solution10,11, which makes the signal proportional to the target concentration. Kelley, et al. have also used a masking DNA template was used to remove false positives and effectively quantify mutated circulating nucleic acids at 0.01% relative to wild-type, rivaling digital PCR 10,11. Thus, a variety of high performance electrochemical methods have been developed for oligonucleotide quantification.

To understand these systems and promote further improvements, researchers have developed useful models to describe the changes in electron-transfer rates through targetinduced conformational changes at electrodes12–15. Huang and White showed that a change in the distance between the redox moiety and the surface could either change the redox moiety from diffusion-limited to adsorbed behavior, or vice-versa13. In the absence of conformational changes, the rate of collisions of the redox moiety with the electrode surface is altered15. Under this mechanism, some real-time electrochemical sensors have been developed using structure switching aptamers that bind proteins 16–18. These assays have also exhibited high performance and have gained popularity in recent years. However, a limitation of these mechanisms is that the target, upon binding, should result in effective change in the electron-transfer rate, which makes assay development nontrivial. Temperature is one important parameter that has received less attention in experimental and theoretical studies of such DNA-based electrochemical biosensors. Slinker and coworkers highlighted this importance in their demonstration that DNA melting could be interrogated at electrode surfaces by square-wave voltammetry (SWV), and they revealed an additional feature of generally increased SWV current at higher temperatures 19. Later, the Plaxco group effectively discriminated single-nucleotide polymorphisms using electrochemical melt curve analysis20. Building upon these works along with our own studies of signal and background in homogeneous fluorescence assays 21,22, here we use a customized thermally


Page 2 of 9

Page 3 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry controlled electrochemical cell to systematically evaluate the effects of temperature, SWV frequency, and DNA hybridization energies in a novel, single-branched proximity assay for DNA at electrode surfaces. We demonstrate that high temperature not only weakens the background complexes in targetinduced DNA hybridization on the surface, but it also contributes to the appreciation of the analytical signal by increasing electron transfer rates. This approach at amplifying electrochemical signal while reducing target-independent DNA hybridization using temperature control should be applicable to a variety of other systems utilizing electrochemical-DNA (EDNA) or aptamer sensors (aptasensors). The study also highlights the key role for assay temperature and suggests that it should not be overlooked when designing such systems.

EXPERIMENTAL SECTION Chemicals and Materials. All solutions were prepared with deionized, ultra-filtered water (Fisher Scientific). The following reagents were used as received: 4-(2-hydroxyethyl)-1piperazineethanesulfonic acid (HEPES) and sodium perchlorate from Alfa Aesar; tris-(2-carboxyethyl) phosphine hydrochloride (TCEP), mercaptohexanol, gold etchant, and chromium etchant from Sigma-Aldrich. Methylene blue (MB)conjugated DNA was purchased from Biosearch Technologies (Novato, CA), purified by RP-HPLC. Thiolated DNAs were obtained from Integrated DNA Technologies (IDT; Coralville, Iowa), with purity confirmed by mass spectrometry. Goldsputtered glass slides (100 nm Au with 5 nm Cr adhesion layer) were purchased from Deposition Research Lab (St. Charles, MO), with dimensions of 1.0 in × 3.0 in × 0.44 in (width, length, thickness). AZ 40XT (positive photoresist) and AZ 300 MIF developer was obtained from MicroChem. Polydimethylsiloxane (PDMS) was purchased from Dow Corning, and dimethyl sulfoxide (DMSO) was purchased from VWR/Anachemia. Preparation of the gold electrode and electrochemical cell. Electrode masks were designed in Adobe Illustrator, and files were sent to Fineline Imaging (Colorado Springs, Colorado) for printing of positive photomasks. The mask design is shown in Figure S-1, where in one glass slide three chips can be made from 1 in2 (645 mm2). Using AZ 40XT photoresist, standard photolithographic procedure was followed to make the photoresist pattern on the gold on glass slide (GoG). Then the GoG was introduced into gold enchant followed by chromium etchant for 30 s and 15 s, respectively. In this way, gold and chromium unblocked by photoresist were removed, resulting in gold electrodes defined by the mask design. Heating this electrode-patterned GoG in DMSO at 110 °C for 30 min removed the positive photoresist from above the gold. The GoG electrodes were rinsed with deionized water followed by ethanol and dried with nitrogen. SWV frequency study at varying temperature. Electrochemical measurements were performed using a Gamry Reference 600 potentiostat. Once the electrodes were immobilized with the desired thiolated-DNA sequences by the above mentioned procedure, 100 nM of the target (varying sequence) and 100 nM of MB-DNA in buffer (10 mM HEPES, 0.5 M NaClO4, pH 7.4) was introduced and incubated overnight at 4 °C. For background measurement, only 100 nM MB-DNA was introduced. The overnight incubation ensured that hybridization reached equilibrium and avoided kinetic effects, and the low temperature minimized evaporation. Following incubation, the chip was placed on our in-house built, custom tem-

perature control system (Figures S-8 and S-9) with the Peltier device held at 4 °C, and the temperature was slowly increased to 15 °C. Once the system reached 15 °C, the platinum counter electrode (CH instruments) and silver/silver chloride (3 M KCl) reference electrode (BASI) were introduced into the reservoir, and the SWV experiment with eighteen different frequencies (4 Hz – 900 Hz) was carried out at 5 °C temperature intervals until 70 °C was reached. All reported temperatures were measured at the aluminum block above the Peltier device, and thermal grease was used between this block and chip to minimize the interfacial gradient. Based on the larger thermal conductivities of the grease, glass, and gold electrode, we assume this measurement to be an accurate representation of the electrode temperature. For all SWV measurements, the potentiostat was programmed to scan the voltage from -0.450 V to 0.00 V with a step size of 1 mV, a pulse height of 50 mV, and a SWV frequency range from 4 to 900 Hz. Single-branched DNA structure. Our previous work on development of the electrochemical proximity assay ECPA for protein quantification24 used target induced DNA-DNA binding stabilization on the surface of the electrode. In this work, we modified the proximity-based ECPA concept and extended it to DNA quantification, in which the target DNA (21 nucleotides) cooperatively binds with a thiolated-DNA and methylene blue-DNA (MB-DNA) to form a three part, singlebranched DNA hybridization structure (Figure 1). The stability of the thiolated-DNA and MB-DNA hybridization (green region) is increased by the presence of target due to the proximity effect, bringing electrochemically active MB-DNA near the electrode in proportion to the amount of target DNA. We

Figure 1. Single-branched cooperative DNA quantification structure. To elucidate the hybridization arrangement(s) giving optimal signal and minimized background, five different thiolatedDNAs were used (green region; n = 7, 9, 10, 12, 14), each with five different target DNAs tested (red region; m = 6, 7, 8, 9, 10), resulting in 25 combinations of binding energies and locations. The blue region was varied to maintain the target DNA at 21 base pairs (s + m = 21). Black regions show structural portions of the sequences not involved in binding.

used five different thiolated-DNA strands (green region; n = 7, 9, 10, 12, or 14) which hybridized at different affinities with MB-DNA. Additionally, for each n, five different oligonucleotide target stabilities were studied (red region; m = 6, 7, 8, 9, 10). In all complexes, the target was maintained as a 21-nt DNA strand by varying the length of the thiolated-DNA binding region (blue region, s) such that s + m = 21. Data analysis. Peak Height: Each set of raw data from square-wave voltametry (including Vstep and Idiff data) was transferred to Microsoft Excel, and a nineteen-point moving average was applied to reduce environmental noise. Following this, a third-order polynomial baseline was calculated near the redox potential of MB-DNA. To do so, the “Linest” function in Excel was used, and data points from -0.400 V to -0.370 V



and -0.15 V to -0.09 V were selected for baseline fitting. The resultant baseline function was subtracted from the 19-point averaged data to get a baseline corrected SWV voltammagram. The maximum current from this graph was used asthe peak height of each particular run. Signal-tobackground differences and ratios were also calculated using peak heights obtained in this way. An example of 19-point averaged data and the calculated baseline is shown in Figure S-3. Electrochemical kinetics: The electrochemical kinetics of the complex at a particular temperature was calculated through the method of Komorsky-Lovric, et al.25. Peak heights at eighteen different SWV frequencies (4 Hz – 900 Hz) were obtained as above, then the peak height was divided by its respective frequency and normalized to the maximum value. This result was plotted as a function of inverse frequency, and the maximum in this plot thereby defined the critical electrochemical reaction time for each experimental condition, giving a measure of the electron transfer kinetics that was independent of hybridization energy. Heat maps: To compare sets of data over a wide range of conditions, two dimensional heat map images were generated from the peak heights or normalized kinetic data using ImageJ software. The “fire” look-up table was used to display data, and heat map color legends were included in each figure. Each individual heat map consisted of 90 by 72 pixels of data (6480 total measurements). Assay calibrations: Assay responses were fit to a five-point logistic regression shown in Equation 1 below,

1

where H was the peak height; C was the concentration of the

target; and a, b, c, and d were fitting constants. The cure fitting was done in Excel using the “Solver” add-in. To calculate a statistically robust lower detection limit (LOD) for nonlinear responses, we followed the method described by Holstein et al.26. By this method, the LOD with 95% confidence interval was obtained. This was achieved by not only considering the standard deviation of the blank but also the deviation of all the sample concentrations.

RESULT AND DISCUSSION The core target recognition concept presented in this work is the novel single-branched, cooperative structure designed for short nucleic acid quantification, as depicted in Figure 1. This form of increased stability by a connector DNA strand was studied in our previous works by fluorescence, electrochemistry, and surface-plasmon resonance27–29. In the present work, this stability increase leads to an increase in MB-DNA molecules at the electrode surface, and SWV current thus increases in proportion to the concentration of the target DNA, resulting in an effective short oligonucleotide quantification method. This study focuses on a thorough characterization of the performance of our assay as a function of hybridization energies, temperature, and SWV frequency, yet the findings should be relevant for a variety of other E-DNA and aptamer-based electrochemical sensors. Signal comparisons with varying complex stability. It is known that for an adsorbed redox reporter, an increase in SWV frequency results in an increase in peak height of SWV current30. For fixed positions of stably hybridized MB-DNA, this effect was observed in all complexes analyzed herein. An example where n = 14, m = 10, and T = 25 °C is shown in Figure S-4. This effect was observed in all cases where complex stability was expected, confirming that the labeled DNA

Figure 2. Heat maps displaying peak height (color intensity) as a function both temperature (y-axis) and SWV frequency (x-axis). This display aids in visualizing the results of the 25 different complexes studied in this work and helps to determine the most optimal conditions for the assay format. For clarity, truncated complexes are shown above and to the right, although all studies included the full complexes as shown in Figure 1.


Page 4 of 9

Page 5 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry strands were being held at the electrode surface. This effect was also observed in heat maps of peak height as a function of both temperature and SWV frequency (Figure 2). Furthermore, from these heat maps we observed that signal increased with increasing numbers of base pairs between MB-DNA and thio-DNA (right to left, n = 7-14) and with increasing pairs between MB-DNA and the target DNA (bottom to top, no sample and m = 6-10), as expected. This increase in stability was also witnessed from the melting temperature (Tm) changes. For n = 9 with no target, the DNA dissociated/melted and signal was lost above 30 °C, whereas for n = 9 and m = 10 the structure was stable through nearly 50 °C. The n = 14 system, even with no target, was stable on the surface up to 70 °C. The heat maps of peak height versus temperature and SWV frequency in Figure 2 highlight the importance of fine-tuning the DNA binding energies in such assays, particularly when developing a system to quantify small oligonucleotide targets. For example, signal was significantly lower in the n = 7 case compared to the n = 14 case, however, there was essentially no difference between the background (no target) and the signal traces in the n = 14 case. The complex with n = 7, while perhaps non-optimal, would function as a sensor, yet the complex with n = 14 would not be useful for target detection. These concepts are explored further in the signal-to-background difference and ratio studies that follow. Moreover, the heat maps illustrate the importance of controlling not only SWV frequency, an easily controlled but sometimes overlooked parameter, but also assay temperature, an often overlooked parameter that requires more specialized analytical systems. Temperature dependence of electrochemical kinetics. As noted above, electrochemical kinetics of the complex at a particular temperature was determined following the method of Komorsky-Lovric et al.25, namely by plotting normalized peak height over frequency (Ip/fswv) versus inverse frequency (1/fswv) to arrive at the critical electrochemical reaction time from the plot maximum. Heat maps of Ip/fswv versus temperature and inverse frequency (1/fswv) using the 25 different complex structures are shown in Figure S-5. It is instructive to note that the critical times (brightest points) were approximately 0.1 s at 15 °C for all complexes, showing that complex stability has a negligible influence over the electron transfer kinetics between MB-DNA and the electrode. Furthermore, as the temperature was increased, all complexes underwent an increase in kinetics (decrease in critical time), observed as an left-upward diagonal shift in the heat maps (Figure S-5). The weakest complexes lost signal at higher temperatures due to melting from the surface. Figure 3 shows snapshots of these electron transfer kinetic effects for four different complexes. The most stable complex is shown in the top set of curves (n = 14, m = 10; with target DNA). As the temperature was increased from 15 to 70 °C, the critical electrochemical reaction time shifted (toward the left) by nearly two orders of magnitude, from 0.3 to 0.01 s. Thus, as temperature was increased, the electrochemical reaction was enhanced. A similar effect was observed in the absence of target DNA (n = 14, no target), except that partial complex melting made the kinetics measurements difficult at 70 °C. Although overall SWV currents were certainly lower for less stable complexes, the electrochemical reaction kinetics followed essentially the same trend with n = 7, m = 10 (with target DNA). However, in the absence of target (n = 7, no target), the complex melting prevented the reaction kinetics

measurements from being made at higher temperatures. These results again confirmed that the complex stability due to DNA hybridization energies had negligible effect on its electron transfer kinetics, which is instead defined largely by the positioning of the MB-DNA with respect to the electrode surface. This temperature enhancement of the electrochemical kinetics can be understood by considering the diffusion layer thickness13. Figure S-6 compares the peak height versus scan rate of a stable complex (n = 14, m = 10) at 15 °C and 40 °C, in which the increase in diffusion layer thickness with temperature is obvious. The linear dependence on scan rate was observed at higher scan rates at 40 °C compared to 15 °C, indi-

Figure 3. Electron transfer kinetics as a function of temperature for selected complexes. Shifts in electrochemical critical times as a function of temperature can be visualized using normalized Ip/fSWV displayed as a function of inverse SWV frequency (1/fSWV), where the maximum represents the critical electron transfer time (in seconds). Similar kinetic trends were observed as temperature was increased, independent of complex stability on the surface (stable hybridization at n = 14; weak hybridization at n = 7). The weakest complex (n = 7, no target) prevented kinetic measurements simply because the complex dissociated/melted from the surface.

cating behavior that more closely resembled an adsorbed molecule, i.e. a molecule confined within the electrochemical diffusion layer. Thus, the increased temperature



Page 6 of 9

Figure 4. Heat maps displaying signal peak height minus background peak height (signal-to-background difference; depicted as color intensity) as a function both temperature (y-axis) and fSWV (x-axis). Excluding the leftmost maps which showed overwhelming background, signals showed a complex relationship on m, n, temperature, and fSWV, suggesting that careful evaluation of these parameters may be necessary in surface-confined, DNA-driven electrochemical assays. Conditions identified for further study are labeled as regions (1) through (4).

Figure 5. Heat maps displaying signal peak height divided by background peak height (signal-to-background ratio; depicted as color intensity) as a function both temperature (y-axis) and fSWV (x-axis). Again, signals showed a complex relationship on m, n, temperature, and fSWV, but the ratio was clearly optimal at low background stabilities (n = 7). The highest ratio was observed at n = 7, m = 10, fSWV = 100 Hz, and T = 30 °C; this region was labeled as (3) in the maps.


Page 7 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry extended the electrochemical diffusion layer further from the electrode, permitting higher electrochemical reaction rates to be observed. These results imply that careful tuning of the complex stabilities and the temperature could give significant enhancements to signal in a variety of surface-confined, DNA driven assay formats using SWV readout. Characterizing signal and background. In this work, we have developed a direct-readout (amplification free) DNA quantification method. Using a novel single-branched hybridization structure that exploits the proximity effect, we compared signal and background in twenty-five different complexes in different hybridization arrangements. Below, we present studies of these complexes from two different perspectives: 1) signal-to-background difference, and 2) signal-tobackground ratio. Signal-to-background difference. Heat maps of signal-tobackground differences of all 25 complexes are shown in Figure 4 as a function of both temperature and fSWV. The addition of target DNA into each complex increased the peak height in all cases except where n = 14. In fact, zero or negative changes

were observed for this complex (blue or black pixels denote negative changes). These effects can be explained by steric hindrances, described by Mahshid et al.31,32 and by the increased stability of background when n = 14. Since the system was designed for “signal-on” readout, the n = 14 complex is ignored in further discussion in this section. Examining all other complexes, we observed a convoluted dependence of signal-to-background differences on m, n, T, and fSWV, where different optimal combinations gave the highest signals for each complex. In fact, we expect that careful evaluation of these parameters may be necessary in any surface-confined, DNA-driven electrochemical assays. From Figure 4, high differences were observed when n = 9 - 12, and two optimal sets of parameters appeared to stand out as having the highest differences (yellow/white spots). The region labeled (1) in the figure has n = 12, m = 9, fSWV ≥ 400 Hz, and T = 45 °C. The second region, labeled (2) is the case where n = 9, m = 8, fSWV ≥ 400 Hz, and T = 35 °C. Other labeled regions, (3) and (4), are discussed below.

Figure 6. Single- and two-step quantification of target DNA with the selected single-branched complexes and condition sets from the heat maps above, (1) through (4). The left side with blue background is for a single-step calibration workflow, whereas the right side with red background is for a two-step calibration workflow. The red curves fitted to all data are four-point logistic curve fits. The standard deviation is presented for all cases (n = 6 electrodes).



Signal-to-background ratio. Viewing these data through a second perspective, signal-to-background ratio, we attained heat maps with different optimal regions (Figure 5). The introduction of sample resulted in the best ratios where background complexes were the weakest, i.e. when n = 7. While the case of n = 12 exhibited high signal-to-background differences (Figure 4), this complex showed very low ratios (Figure 5). For n = 7, which has the best ratios, as m was increased the ratio increased, giving the optimal parameters as n = 7, m = 10, fSWV = 100 Hz, and T = 30 °C. This region was marked as (3) in Figures 4 and 5. Difference and Ratio Comparisons. On a composite view of signal-to-background differences and ratios, the n = 9 complex was found to be unique, giving moderate signals in both cases. Specifically, region (2) in Figures 4 and 5 (n = 9, m = 8, fSWV ≥ 400 Hz, and T = 35 °C) can be selected as a favorable complex for DNA quantification considering this joint viewpoint. On the other hand, all of these conclusions are based on mixand-read workflow, or a single-step method. If we consider that a heterogeneous workflow can be used for stepwise introduction of target, the binding between target and thiolatedDNA can be stronger without worrying about excessive background formation. Data from the selected four sets of conditions were extracted and overlapped in Figure S-7 in two dimensions. The left and right figures compare the signal-to-background differences and ratios, respectively, while the top figures are plotted versus frequency and the bottom as a function of temperature. The four selected set of conditions are highlighted with asterisks. This demonstrates our previous conclusions that (1) gave the maximum difference (purple curves), (2) showed good signal and a moderate ratio (green curves), (3) showed a very high ratio (red curves), and (4) is optimal for low-temperature detection (blue curves). Oligonucleotide quantification. With this information in hand, assay calibration experiments were carried out under the four sets of selected conditions, with results shown in Figure 6. These data compare the calibration curves of single- and two-step workflows for all conditions. Condition (1) did not show an appreciable increase in peak height induced by target through 50 nM concentration, in both the single- and two-step workflows. This is likely due to the very strong background formation through binding between MB-DNA and thio-DNA. We observed a similar response with n = 14 in our previous experiments. The other three conditions showed a sigmoidal relationship between the SWV peak height and target DNA concentrations. These curves were fitted with a four-point logistic equation, shown as red curves in the figures. In condition (2) high signal and an increasing response was observed in the low nanomolar region, and the single-step workflow exhibited higher sensitivity and dynamic range compared to the two-step workflow. Similar trends were observed in condition (3), with slightly improved sensitivity at low concentration but lower overall signal. In contrast, the low temperature assay condition (4) exhibited better results with the two-step workflow. The limits of detection (LOD) of these assay formats were all between 2 and 20 nM. Under the three responsive condition sets, (2) through (4), LODs were calculated and compared as shown in Figure 7, where blue and red bars represent the single-step and two-step workflows, respectively. It was observed that condition (2) was moderate in both the cases, while having two- to three-fold more sensitivity (see Figure 6) in

comparison to the other two conditions. Condition (3) had lower LOD with the one-step workflow yet showed the worst LOD in the two-step method. By contrast, condition (4) had a smaller LOD using a two-step protocol compared to the single-step approach. Thus, for a two-step workflow, one should choose condition (4); for a single-step workflow, condition (3) is optimal; and the highest sensitivity and widest range can be achieved with condition (2). Stated differently, for single-step workflows, a balanced increase in both m and n can provide high sensitivity, while maximizing m and minimizing n can result in lower assay LOD. For two-step workflows, one should increase the stability between the sample and capture probe on the surface while minimizing target-independent background formation. More generally, this study has shown that careful control over surface hybridization energies, temperature, and SWV frequency can provide significant flexibility in operating our novel, single-branched electrochemical DNA assay platform.

CONCLUSIONS In this work, we have presented a new assay design while also developing a generalizable approach for fine-tuning or optimizing electrochemical DNA sensors. We carefully controlled hybridization energies, square-wave voltammetry frequency, and temperature. Results demonstrate that temperature is indeed a very important factor in optimizing such assays. In homogeneous, DNA-driven assays with optical readout, such as pincer assays33–35 or proximity ligation assays 36, our group and others have shown that precise temperature control can serve to reduce background hybridizations or ligations, yet fluorescence quantum yield tends to decrease as temperature is increased. By contrast, we show here that with precise temperature control in DNA-driven electrochemical assays, electron transfer rates can be increased while background hybridizations are simultaneously minimized. This work highlights the importance of understanding such effects in our novel singlebranched DNA assay, and similar trends will likely be observed in many electrochemical DNA assays and aptasensors. Regarding our single-branched proximity assay developed here, the precise control of the stated parameters revealed an inherent flexibility in the assay. Various formats of the assay could be chosen for low temperature (15 °C) readout, higher temperature (45 °C) readout, minimized LOD, or maximized sensitivity and range. While some reported DNA- or aptamerdriven electrochemical assays have optimized SWV frequen-

Figure 7. Limit of detection comparison for the three sets of conditions with responsive assays, (2) through (4). Results show the important effects of binding energy, assay temperature, and workflow/procedure on LOD.


Page 8 of 9

Page 9 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry cy, very few have simultaneous studied the temperaturedependent effects as shown here. Hence, a straightforward future direction is to employ a temperature-controlled system as shown here, generate assay response heat maps, and optimize the temperature and frequency dependences of these other assays.

(15) (16)

(17)

ASSOCIATED CONTENT Supporting Information (SI) is available free of charge on the ACS Publications website. Experimental details and supporting figures and tables.

(18)

(19)

AUTHOR INFORMATION (20)

Corresponding Author * Prof. Christopher J. Easley, [email protected]

ACKNOWLEDGMENT

(21)

Funding for this work was provided by the National Institutes of Health through award R01 DK093810, and by the National Science Foundation through award CBET-1403495.

(22)

REFERENCES

(23)

(1)

(2) (3) (4) (5) (6) (7) (8)

(9)

(10) (11) (12) (13) (14)

Flechsig, G.-U.; Peter, J.; Hartwich, G.; Joseph Wang, § and; Gründler, P. Langmuir 2005, 21 (17), 7848– 7853. Shen, Z.; Sintim, H. O.; Semancik, S. Anal. Chim. Acta 2015, 853, 265–270. Peterlinz, K. A.; Georgiadis, R. M. J. Am. Chem. Soc 1997, 119 (14), 3401–3402. Fan, C.; Plaxco, K. W.; Heeger, A. J. Proc. Natl. Acad. Sci. U. S. A. 2003, 100 (16), 9134–9137. Shen, Z.; Nakayama, S.; Semancik, S.; Sintim, H. O. Chem. Commun. 2012, 48 (61), 7580. Lai, R. Y. In Methods in Enzymology; 2017; Vol. 589, pp 221–252. Idili, A.; Amodio, A.; Vidonis, M.; FeinbergSomerson, J.; Castronovo, M.; Ricci, F. . Lai, R. Y.; Lagally, E. T.; Lee, S.-H.; Soh, H. T.; Plaxco, K. W.; Heeger, A. J. Proc. Natl. Acad. Sci. U. S. A. 2006, 103 (11), 4017–4021. Patterson, A.; Caprio, F.; Vallée-Bélisle, A.; Moscone, D.; Plaxco, K. W.; Palleschi, G.; Ricci, F. Anal. Chem. 2010, 82 (21), 9109–9115. Das, J.; Ivanov, I.; Sargent, E. H.; Kelley, S. O. J. Am. Chem. Soc. 2016, 138 (34), 11009–11016. Das, J.; Ivanov, I.; Montermini, L.; Rak, J.; Sargent, E. H.; Kelley, S. O. Nat. Chem. 2015, 7 (7), 569–575. White, R. J.; Rowe, A. A.; Plaxco, K. W. Analyst 2010, 135 (3), 589–594. Huang, K.-C.; White, R. J. J. Am. Chem. Soc. 2013, 135 (34), 12808–12817. Uzawa, T.; Cheng, R. R.; White, R. J.; Makarov, D. E.; Plaxco, K. W. J. Am. Chem. Soc. 2010, 132 (45), 16120–16126.

(24)

(25) (26) (27)

(28) (29) (30) (31) (32) (33) (34) (35) (36)

White, R. J.; Plaxco, K. W. Anal. Chem. 2010, 82 (1), 73–76. Mage, P. L.; Ferguson, B. S.; Maliniak, D.; Ploense, K. L.; Kippin, T. E.; Soh, H. T. Nat. Biomed. Eng. 2017, 1 (5), 70. Zhou, Q.; Kwa, T.; Gao, Y.; Liu, Y.; Rahimian, A.; Revzin, A. Lab Chip 2014, 14 (2), 276–279. Arroyo-Currás, N.; Somerson, J.; Vieira, P. A.; Ploense, K. L.; Kippin, T. E.; Plaxco, K. W. Proc. Natl. Acad. Sci. U. S. A. 2017, 114 (4), 645–650. Wohlgamuth, C. H.; McWilliams, M. A.; Slinker, J. D. Anal. Chem. 2013, 85 (3), 1462–1467. Yang, A. H. J.; Hsieh, K.; Patterson, A. S.; Ferguson, B. S.; Eisenstein, M.; Plaxco, K. W.; Soh, H. T. Angew. Chemie Int. Ed. 2014, 53 (12), 3163–3167. Kim, J.; Hu, J.; Bezerra, A. B.; Holtan, M. D.; Brooks, J. C.; Easley, C. J. Anal. Chem. 2015, 87 (19), 9576– 9579. Hu, J.; Easley, C. J. Anal. Chem. 2017, 89 (16), 8517– 8523. Brooks, J. C.; Ford, K. I.; Holder, D. H.; Holtan, M. D.; Easley, C. J. Analyst 2016, 141 (20), 5714–5721. Hu, J.; Yu, Y.; Brooks, J. C.; Godwin, L. A.; Somasundaram, S.; Torabinejad, F.; Kim, J.; Shannon, C.; Easley, C. J. J. Am. Chem. Soc. 2014, 136 (23), 8467–8474. Komorsky-Lovrić, Š.; Lovrić, M. Anal. Chim. Acta 1995, 305 (1–3), 248–255. Holstein, C. A.; Griffin, M.; Hong, J.; Sampson, P. D. Anal. Chem. 2015, 87 (19), 9795–9801. Arugula, M.; Yang, F.; Somasundaram, S.; Easley, C. J.; Shannon, C.; Simonian, A. ECS Trans. 2015, 66 (36), 19–29. Hu, J.; Wang, T.; Kim, J.; Shannon, C.; Easley, C. J. J. Am. Chem. Soc. 2012, 134 (16), 7066–7072. Kim, J.; Hu, J.; Sollie, R. S.; Easley, C. J. Anal. Chem. 2010, 82 (16), 6976–6982. Osteryoung, J. G.; Osteryoung, R. A. Anal. Chem. 1985, 57 (1), 101A–110A. Mahshid, S. S.; Camiré, S.; Ricci, F.; Vallée-Bélisle, A. J. Am. Chem. Soc. 2015, 137 (50), 15596–15599. Mahshid, S. S.; Vallée-Bélisle, A.; Kelley, S. O. Anal. Chem. 2017, 89 (18), 9751–9757. Heyduk, E.; Heyduk, T. Anal. Chem. 2005, 77 (4), 1147–1156. Heyduk, E.; Dummit, B.; Chang, Y.-H.; Heyduk, T. Anal. Chem. 2008, 80 (13), 5152–5159. Heyduk, T. Biophys. Chem. 2010, 151 (3), 91–95. Fredriksson, S.; Gullberg, M.; Jarvius, J.; Olsson, C.; Pietras, K.; Gústafsdóttir, S. M.; Östman, A.; Landegren, U. Nat. Biotechnol. 2002, 20 (5), 473– 477.


Understanding signal and background in a ... - ACS Publications

Recommend Documents