16163028, 2022, 37, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202201881 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [14/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Research Article www.afm-journal.de A Programmable Electrochemical Y-Shaped DNA Scaffold Sensor for the Single-Step Detection of Antibodies and Proteins in Untreated Biological Fluids Andrea Idili, Andrea Bonini, Claudio Parolo, Ruslán Alvarez-Diduk, Fabio Di Francesco, and Arben Merkoçi* enable mass serological screening directly Proteins and antibodies are key biomarkers for diagnosing and monitoring from untreated clinical samples, thereby specific medical conditions. Currently, gold standard techniques used for reducing the time lag between the diag- their quantification require laborious multi-step procedures, involving high nosis and the treatment (e.g., from infec- costs and slow response times. It is possible to overcome these limitations tive to autoimmune diseases),[3,4] enabling easier global health monitoring and sur- by exploiting the chemistry and programmability of DNA to design a veillance (i.e., monitoring vaccination sta- reagentless electrochemical sensing platform. Specifically, three DNA tuses).[5–7] This new diagnostic technology single strands are engineered that can self-assemble into a Y-shaped DNA needs not only to be rapid (i.e., taking nanostructure that resembles one of the IgGs. In order to convert this <10 min), quantitative, highly specific, DNA nanostructure into a responsive DNA-scaffold bioreceptor, it is and adaptable to a wide range of diseases, but also to have a high-throughput (i.e., modified including two recognition elements, two redox tag molecules, capable of precisely analyzing hundreds of and a thiol group. In the absence of the target, the scaffold receptor samples in a short time) and be easy to use can efficiently collide with the electrode surface and generate a strong so that even non-trained personnel would electrochemical signal. The presence of the target induces its bivalent be able to perform the tests properly.[8] In binding, which produces steric hindrance interactions that limit the contrast, current clinical approaches do receptor’s collisional activity. In its bound state, the redox tags can therefore not fulfill all these analytical requirements. For example, laboratory-based techniques approach the surface at a slower rate, leading to a signal decrease that is such as the enzyme-linked immuno- quantitatively related to the target concentration. The Y-shape DNA scaffold sorbent assay (ELISA) and Western blot sensor can detect nanomolar concentrations of antibodies and proteins in assay require reagent-intensive and multi- <15 min with a single-step procedure directly in untreated biological fluids. step processes (e.g., washings, incuba- tions, and sample treatments).[9,10] On the other hand, point-of-care devices, such 1. Introduction as lateral flow assays (LFA), can perform the naked eye detec- tion of specific biomarkers in <15 min, however, they show low The development of a sensing platform capable of supporting sensitivity and specificity, as well as the inability to precisely the single-step detection of clinically relevant antibodies and pro- quantify the target.[3,11,12] Consequently, a sensing technology teins would strongly impact the current diagnostic approaches that synergically combines the high sensitivity and specificity of through high-frequency tests.[1,2] This, for example, would laboratory-based techniques with the speed and convenience of A. Idili, A. Bonini, C. Parolo, R. Alvarez-Diduk, A. Merkoçi A. Bonini, F. Di Francesco Catalan Institute of Nanoscience and Nanotechnology (ICN2) Department of Chemistry and Industrial Chemistry campus UAB, Bellaterra, 08193 Barcelona, Spain University of Pisa E-mail: arben.merkoci@icn2.cat 56124 Pisa, Italy C. Parolo The ORCID identification number(s) for the author(s) of this article ISGlobal, Barcelona Centre for International Health Research (CRESIB) can be found under https://doi.org/10.1002/adfm.202201881. Hospital Clínic (Department of International Health) Universitat de Barcelona © 2022 The Authors. Advanced Functional Materials published by Wiley-VCH 08036 Barcelona, Spain GmbH. This is an open access article under the terms of the Creative Com- A. Merkoçi mons Attribution-NonCommercial-NoDerivs License, which permits use CSIC and the Barcelona Institute of Science and Technology (BIST) and distribution in any medium, provided the original work is properly cited, 08036 Barcelona, Spain the use is non-commercial and no modifications or adaptations are made. A. Merkoçi Institució Catalana de Recerca i Estudis Avançats (ICREA) DOI: 10.1002/adfm.202201881 08010 Barcelona, Spain Adv. Funct. Mater. 2022, 32, 2201881 2201881 (1 of 11) © 2022 The Authors. Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 2022, 37, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202201881 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [14/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.afm-journal.de Figure 1. Y-shaped DNA scaffold sensors consists of a nucleic acid “scaffold” nanostructure designed i) to mimic the natural shape of the bivalent immunoglobulin G antibodies and ii) to be bound to a gold surface via a flexible hexanethiol linker. The arms of the scaffold are modified with two recognition molecules and two redox reporters (here methylene blue). (Left) Binding of the target antibody to the recognition elements reduces the frequency with which the redox reporters collide with the electrode surface, resulting in a decrease of the faradic current collected using square wave voltammetry. (Right) Representative square wave voltammograms of the unbound and target-bound scaffold sensor (for the detection of 100 × 10−9 m anti-digoxigenin antibody in 100% serum). point-of-care testing could have a significant impact on current inspiration from the structure of natural Immunoglobulin diagnostics. G (IgG) antibodies (Figure 1). Specifically, we employed as a New biosensing platforms have been developed over the last recognition element a DNA “scaffold” nanostructure ration- two decades to achieve reagentless, single-step detection of anti- ally designed to assume a Y-shaped structure through the self- bodies (or proteins) directly in untreated biological fluids.[13–15] assembly of three modified DNA single strands (Figure 1). These platforms exploit synthetic, redox-labeled nucleic acid The arms of the nanostructure are modified with two recogni- molecules as responsive recognition elements that generate elec- tion elements, placed at a distance equal to the two paratopes trochemical signals in response to the presence of the selected of human IgGs, and two redox tag molecules (i.e., methylene target.[16–18] For example, biosensors based on DNA aptamers blue). Thanks to this configuration, the scaffold sensor can or antigen-tagged DNA probes can fold/unfold in the presence bind bivalently to its target inducing stronger steric hindrance of the biomolecular target (i.e., they undergo a binding-induced interactions and the suppression of the faradic current. At the conformational change) generating a specific electrochemical same time, the use of two methylene blue reporters generates signal.[19–21] Besides their elegant design and versatility, their an electrochemical signal strong enough to support the detec- main advantage is that they exhibit higher specificity, since to tion of the target directly in biological fluids. Using three dif- generate the signal, the bioreceptor has to undergo a binding- ferent recognition elements (digoxigenin, DNP, and biotin), we induced conformational change; an approach similar to the one showed how our sensing platform is highly programmable and observed in natural receptors.[22,23] However, these biosensors can fully support the rapid, single-step detection of antibodies are often expensive (due to internal modifications[24] of the DNA and proteins directly in undiluted biological matrices (serum sequence), and require a precise thermodynamic characteriza- and artificial saliva) at nanomolar concentrations. tion to enhance the conformational change on the electrode sur- face, which directly affects the overall analytical performance of the sensor.[25,26] Sensors based on DNA-scaffold receptors (i.e., 2. Results and Discussion double-stranded DNA presenting a recognition element) have been deployed to fully succeed in the detection and monitoring 2.1. Design (for many hours) of antibodies and proteins in undiluted samples (e.g., serum, blood, and saliva).[27–32] Their advantages rely on a Our strategy to develop a programmable DNA-scaffold sensor minimalist design since the signal production depends mostly for the rapid, single-step detection of antibodies (or proteins) is on the steric hindrance produced by the binding of the antibody based on the use of three complementary DNA single strands or protein to its recognition element that is attached to the DNA that have been designed to be assembled in a Y-shaped nano- scaffold. Despite these sensors being able to detect antibodies, structure (Figure 2). The sequence of each single strand has proteins, and even viral particles in a single step, the generation two “stem” portions (20-base; Figure 2A), separated by a poly(T) of the signal is strongly affected by the size of the selected recog- linker of five bases, which have been designed to bind their nition element (epitope or antigen).[30–33] Moreover, their design complementary “stem” sequences present in the other two do not take advantage of the structure of natural receptors such strands (Figure 2A, red, green, and blue domains). To induce as antibodies or multivalent proteins, as any sufficiently large binding with the selected target antibody and to adapt it so monovalent target that induces adequate steric hindrance can that it can support an electrochemical platform, we modified generate a measurable signal. the DNA nanostructure with specific functional groups and In this study, we expand the set of DNA scaffold sensing electroactive tag molecules. More specifically, the first strand architectures by introducing a novel geometry that takes (anchoring strand 1, Figure 2A) was modified at 5′-end with a Adv. Funct. Mater. 2022, 32, 2201881 2201881 (2 of 11) © 2022 The Authors. Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 2022, 37, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202201881 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [14/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.afm-journal.de programmability of DNA as nanomaterial[36–38] to engineer a Y-shaped nanostructure . At the same time, we exploited the random coil conformation of poly(T) linkers[39,40] to intro- duce high flexibility to the two responsive arms, and more in general, to the overall structure.[41,42] This configuration was designed to mimic the high conformational variability and dynamics observed in natural IgGs, which are crucial for their outstanding binding activity.[43–46] We therefore believe that this geometry should guarantee the bivalent binding between one single antibody and one Y-shaped DNA scaffold. We thus validated the design of our scaffold receptor using the nucleic acid folding predictor NUPACK.[47] This software confirmed the correct design and assembly of the three strands. As expected, the three double strand “stem” portions were found to be interconnected by poly(T) linkers forming a Y-shape structure (Figure S1, Supporting Information). After this simulative test, we experimentally demonstrated the assembly of the Y-shaped DNA structure (Figure 2B) using non-denaturing polyacryla- mide gel electrophoresis (PAGE). PAGE is a valid and reliable method for characterizing the formation and yield of DNA nanostructure assemblies.[48–50] This experiment clearly showed that mixing the three modified oligos in the same solution using a 1:1:1 molar ratio, at room temperature leads to the for- mation of a single band with the lowest mobility corresponding Figure 2. A) The Y-shaped scaffold structure is made up of three function- to the assembled scaffold receptor (Figure 2B, lane 6). In con- alized DNA single strands: the “anchoring strand 1” modified at its two ends with a thiol group, and a methylene blue molecule; the “recognition trast, the single strands (Figure 2C, lanes 1,2, and 3) and the two strand” modified with two recognition molecules on its extremities; and self-complementary duplex dimers (Figure 2C, lanes 4 and 5) the “anchoring strand 2” that holds only a methylene blue molecule. Their showed a higher mobility due to their lower weights and self-assembly (driven by Watson-Crick-Franklin interactions) induces the smaller size. These data are in agreement with previous studies formation of three duplex “stem” portions (length of 20 base pairs each; involving the use of similar DNA-based structures based on red, green, and blue domains), interconnected by three linkers composed of a three-way junction.[51,52] After demonstrating the correct five thymines, resulting in a Y-shaped nanostructure. The single strand core formed by the poly(T) linkers ensures that the two recognition arms have assembly, we set out the immobilization and characterization of sufficient flexibility and conformational variability,[41,42] thus mimicking find- our Y-shaped receptor on the electrochemical platform. ings observed with IgG antibodies, where the two paratopes can finely mod- ulate their distance to improve the bivalent binding with their target.[43–46,53] B) The non-denaturing polyacrylamide gel electrophoresis (PAGE) experi- 2.2. Sensing Mechanism ment confirms the formation of the scaffold nanostructure. Lanes 1, 2, and 3 show individual bands of the DNA single strand which, as expected, display Y-shaped DNA-scaffold sensors can promptly detect the pres- the highest mobility, and correspond to the "anchoring strand 1”, “anchoring strand 2”, and “recognition strand”, respectively. Lanes 4 and 5 show indi- ence of the selected target antibody when interrogated using vidual bands corresponding to the hybridization of the two anchoring square-wave voltammetry (SWV) (Figures S2 and S3, Sup- strands and the hybridization of “anchoring strand 2” and the “recognition porting Information). We first demonstrated how the signal- strand”, respectively. Finally, lane 6 shows the band corresponding to the transduction mechanism relies on the collisional activity of the efficient formation of the full Y-shaped DNA scaffold, which, as expected, Y-shaped DNA scaffold receptor with the gold electrode sur- displays the lowest mobility in the gel. face. To achieve this, we selected the detection of anti-digoxi- genin antibodies as a testbed (Figure 3A, left). We decided to thiol group (to anchor the nanostructure to the gold electrode use the steroid digoxigenin as a recognition element because surface) and at the 3′-end with the methylene blue. Both ends this system has been previously exploited to characterize other of the second strand (recognition strand, Figure 2A) were modi- DNA nanotechnology-based serological sensors,[24,33,54,55] due to fied with two equal recognition molecules. Last, only the 5′-end the ease of synthesizing digoxigenin-modified DNA molecules. of the third strand (anchoring strand 2, Figure 2A) was modified In the absence of the target, the modified Y-shaped scaffold with the methylene blue. Once in solution, the three functional- receptor is free to collide with the electrode surface (Figure 1, ized strands hybridize to form a Y-shaped DNA nanostructure left) and can produce a large current at the redox potential where the “stem” portion 1 (blue domain) is connected to the expected for methylene blue (Figure 1, right). The presence of gold electrode surface, while both arms (stem portions 2 and 3, the antibody decreases this faradic current because the bulky red and green domains) are oriented toward the bulk solution antibody-scaffold receptor complex limits the movements that sharing the same methylene blue and recognition molecules. allow the redox reporters to approach the electrode surface. To The design of our scaffold sensor takes advantage of the demonstrate this collisional mechanism and how binding-induced natural shape of IgGs and their bivalent interactions with steric hindrance interactions drive the signal-transduction their molecular targets.[34,35] Specifically, we exploited the mechanism of the sensor, we characterized how the signal gain Adv. Funct. Mater. 2022, 32, 2201881 2201881 (3 of 11) © 2022 The Authors. Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 2022, 37, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202201881 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [14/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.afm-journal.de Figure 3. A) To characterize our new sensing platform, as model system, we detected anti-Dig antibodies using two molecules of digoxigenin (Dig) as recognition elements. We designed four different DNA-scaffold variants introducing in their structures different degrees of flexibility (removing one poly(T) linker from the core; red variant) and positioning the recognition molecules (removing or moving them in the inner part of the structure; black and blue variant). B) Native PAGE gel confirms the full assembly of the scaffold variants. C) The electrochemical detection of anti-digoxigenin antibodies using the four different variants indicates that the one producing the most sensitive response with the biggest signal change is the native variant (green line). The error bars reflect standard deviations derived using at least three independently fabricated sensors. D) Representative square wave voltammograms collected in the absence (gray) and in the presence of the antibody target (colored), for the different variants (for the detection of 100 × 10−9 m anti-digoxigenin antibody in buffer). (i.e., the relative change in signal upon the addition of the satu- the other hand, the amount of receptor chemisorbed on the rating target) is affected by 1) the frequency of the interrogating surface strongly affects the magnitude of the sensor’s signal SWV potential pulse; and 2) the density with which the scaffold gain.[28,33,58] More specifically, at high densities, receptors are receptors are packed on the electrode surface. highly packed on the electrode surface and, as they are in close Previous studies have indicated that the signal gain of DNA- contact with each other, they can limit their collisional move- scaffold sensors exhibits a slight dependence on the SWV ment by decreasing the overall signal change in presence of the frequency applied during the measurements.[56,57] To better target. At low densities, the receptors are far from each other, understand this observation, their sensing mechanism needs and are free to approach the surface resulting in a higher signal to be taken into account. As described above, the binding change in the presence of the antibody target. with an antibody induces sufficient steric hindrance to slow To demonstrate the collisional mechanism of the Y-shaped down the collisional dynamics of the redox tags with the elec- DNA scaffold, we tested our sensors in the presence of satu- trode (without altering the conformational state of the scaf- rating (100 × 10−9 m) anti-DIG antibodies and characterized fold receptor). The binding event does not change the electron the sensor’s response over SWV frequencies ranging from transfer rate of the methylene blue molecules but only reduces 5 to 1000 Hz. We also used different DNA-probe densities by the efficiency of the process. Therefore changes in the fre- changing the concentration of scaffold receptors used during quency of the SWV slightly impact the magnitude and the sign the surface functionalization step (Figure S2, Supporting (i.e., producing either signal-on or signal-off behavior) of the Information). As expected, all the scaffold sensors produced signal change observed.[56,57] Specifically, the sensors display a a signal-off behavior and the magnitude of their signal gain signal-off behavior for almost all the range of SWV frequen- was inversely related to the density of the DNA probe on the cies (from 1 to 1000 Hz) and a comparable signal gain.[56,57] On electrode. The largest signal gain (≈43%) and highest affinity Adv. Funct. Mater. 2022, 32, 2201881 2201881 (4 of 11) © 2022 The Authors. Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 2022, 37, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202201881 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [14/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.afm-journal.de (Figure S3, Supporting Information) for the target were target (Figure 3C,D). We found that the original and modular observed with the lowest scaffold concentration (25 × 10−9 m). variants produce a Langmuir isotherm binding curve, and their On the other hand, sensors fabricated using higher concentra- signal can be related to the amount of target antibody present in tions of DNA scaffold probe displayed lower signal gains and solution (Figure 3A,C; red, green, and blue curves). The native higher dissociation constants (KD) against their target antibody variant produced the largest signal change (up to −42.9 ± 1.1%) (Figure S3, Supporting Information). The overall data therefore and the highest affinity (KD 2.2 ± 0.3 × 10−9 m; Figure 3C, green confirm that the signal-transduction mechanism of our sensor curve). On the other hand, the two modular variants produced is based on a collisional mechanism. a lower signal change (≈−25%) and showed lower affinities (KDs of 20.5 ± 9.5 × 10−9 m and 18.0 ± 1.4 × 10−9 m, for the mod- ular variant and the modular internal variant, respectively) 2.3. Sensor Characterization (Figure 3C, red and blue curves). We believe these data sup- port our initial hypothesis and can be further explained con- The specific architecture of the Y-shaped DNA scaffold receptor sidering two main factors. First, in the more constrained and leads to a superior analytical performance compared to clas- stiffer native variant, the arms have a lower conformational sical DNA scaffold sensors. To demonstrate this, we designed variability (keeping the two epitopes always at a distance, which three new variants in order to induce a floppier structure with facilitates their recognition by the two paratopes), which helps a higher conformational variability. The aim was to enhance the bivalent binding of the antibody (higher affinity). In the the flexibility of the two recognition arms and affect the biva- case of the modular variants, since the arms are free to move, lent binding process with the target antibody (Figure 3A). To they display a higher conformational dynamic that negatively achieve these new variants, we kept the same sequences, and affects the binding with the two paratopes. This could lead to changed the single strand poly(T) core and location of the an alternative binding process where two antibodies bind one recognition elements. Specifically, the first variant, called the single scaffold receptor (lower affinity). Second, because the “modular variant”, has four strands, where the recognition modular variants share recognition arms with a higher flex- strand is substituted by two, non-connected “stem” portions ibility, their high conformational dynamics increases the elec- (Figure 3A, red variant). This architecture allows for a higher tron transfer between the methylene blue molecules and the degree of mobility of the two arms, which inevitably affects electrode surface, leading to a higher background signal and the conformational dynamics of the recognition elements. lower signal gain. Finally, we tested the “control variant” which, The second is the “modular internal variant”, which is like the as expected, did not produce any measurable electrochemical previous one, but with the recognition elements (digoxigenin) signal change for any of the concentrations of antibody used exposed in the inner part of the structure, thus in contact with (up to 300 × 10−9 m; Figure 3C, black curve). The outcome of each other (Figure 3A, blue variant).This architecture mimics these experiments proved that the native variant is indeed the the classic DNA-scaffold sensor where there is only a single rec- one that gives the best analytical performance. ognition element for each probe (i.e., the two epitopes being so The new Y-shaped DNA-based scaffold sensors are sensi- close to each other, only one paratope can bind) while keeping tive and selective. The anti-DIG antibody sensors respond the Y-shaped structure, which facilitates the comparison with sensitively to their target at low-nanomolar concentrations, dis- the native variant.[28] Finally, the third is the “control variant”, playing a useful dynamic range of 0.3 to 18 × 10−9 m (Figure 4A; which has the same structure as the original one, but without green curve). The sensing platform also exhibits specificity, as the digoxigenin molecules, and therefore, it is not able to bind it does not respond to non-targeted antibodies and proteins. To the target (Figure 3A, black variant). Before testing our electro- demonstrate this, we challenged it against anti-DNP antibodies chemical sensors fabricated with the three different scaffold and streptavidin (Figure 4A; red and blue curves). As expected, variants, their assembly was verified by PAGE (Figure 3B, lane no measurable electrochemical signal was observed, not even 5,6, 7, and 8). This electrophoresis experiment showed four in the presence of 300 × 10−9 m of anti-DNP antibodies and distinct bands with a similar mobility corresponding to each streptavidin. This indicates that the response obtained using of the variants. This confirmed that all Y-shaped structures can anti-DIG antibodies is due to the specific recognition of the be assembled with the same efficiency because they share the antigens. To further confirm the multivalent binding between same sequences (Figure 2B). Finally, to demonstrate the dif- the receptor and the target antibody, we used the anti-DIG Fab ferent flexibility/stiffness of the variants, we performed a PAGE fragment as a target. We chose this antibody fragment because using a higher voltage (200 V) (Figure S4, Supporting Informa- it displays a single paratope (i.e., does not exhibit any multiva- tion). Under these conditions, the control and the native variant lent behavior) and lower molecular weight (≈50 kDa reducing (Figure S4 (Supporting Information), lanes 1 and 4) displayed the steric hindrance induced by the binding) than the IgGs the same mobility, which was slightly lower than the modular (≈150 kDa).[59] As expected, the use of the anti-DIG Fab frag- variants (Figure S4 (Supporting Information), lanes 2, and 3). ment produced a lower signal change (−31.7 ± 1.7%) and a This could be due to the higher stiffness of the structures lower affinity (KD 4.8 ± 1.9 × 10−9 m) than those observed with formed only by three strands with respect to the floppier and the complete, bivalent antibody. flexible modular variants formed by four strands. Our sensing platform is not only sensitive and specific, but After confirming the correct assembly of the Y-shaped DNA- also rapid, and enables the single-step detection of the target scaffold receptors, we fabricated the electrochemical sensors in untreated biological fluids. To characterize the sensor’s equi- and tested them by collecting their voltammetric signals with libration time constant, we challenged our sensors using three increasing concentrations of the anti-digoxigenin antibody different concentrations of anti-DIG antibodies: 1, 10, and Adv. Funct. Mater. 2022, 32, 2201881 2201881 (5 of 11) © 2022 The Authors. Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 2022, 37, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202201881 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [14/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.afm-journal.de Figure 4. A) The sensor shows excellent sensitivity and specificity, as it is non-responsive to the presence of anti-DNP antibodies and streptavidin. At the same time, the sensor detects anti-DIG antibodies at low nanomolar concentrations (green curve). To test its performance against monovalent targets, we challenged the sensor against anti-DIG Fab, which produces a lower signal change and sensitivity than those achieved using the full anti- DIG IgG (orange curve). B) The sensor responds rapidly to the presence of anti-DIG antibodies. C) The sensor supports the single-step detection of the target working directly in undiluted matrices including artificial saliva (blue line) and serum (orange line), showing efficient nanomolar detection of the selected target antibody. The error bars reflect standard deviations derived using at least three independently fabricated sensors. D) Representative square wave voltammograms collected in the absence (gray) and in the presence of the antibody target (colored), for the different biological fluids (for the detection of 100 × 10−9 m anti-digoxigenin antibody in serum and artificial saliva). 100 × 10−9 m (Figure 4B). Fitting the binding kinetics to a single comparable performance to the previous system and to previous exponential decay, we estimated a time constant of 56.3 ± 1.7 s studies. [33,60] As expected, the native variant displays higher with 100 × 10−9 m antibody, 11.3 ± 0.4 min with 10 × 10−9 m, analytical performance than the modular variant (Figure S5, and 20.4 ± 3.3 min with 1 × 10−9 m: these values show that the Supporting Information). As previously demonstrated response is sufficiently rapid to use the sensor at the point of care. (Figure 3), this is due to the shape of the target which is a We then characterized the sensor response in undiluted serum bivalent immunoglobulin antibody. Again, the sensors are suf- and artificial saliva (Figure 4C,D, blue curve) spiking increasing ficiently specific to support clinically relevant measurements. concentrations of anti-DIG antibodies from 10 × 10−12 m to They do not respond significantly when challenged with their 300 × 10−9 m. Even in the presence of these undiluted, com- protein interferants (Figure 5A) and they can work perfectly in plex biological matrices, the sensor promptly responded to the undiluted biological fluids, such as serum and artificial saliva. presence of the target in the nanomolar range. More specifi- Specifically, for the DNP system, in artificial saliva the sensor cally, using undiluted artificial saliva, we obtained a maximum produced a maximum signal change of −42.5 ± 3.0% and a KD signal change of −28.8 ± 1.9% and a KD of 12.1 ± 3.4 × 10−9 m of 6.7 ± 2.0 × 10−9 m; while in serum, it produced a maximum (Figure 4C,D, blue curve). When using undiluted serum, we signal change of −28.2 ± 1.6% and a KD of 3.0 ± 0.8 × 10−9 m obtained a maximum signal change of −18.3 ± 0.9% and a KD of (Figure 5B). As observed for the previous system, however, the 3.6 ± 0.9 × 10−9 m (Figure 4C, orange curve). gain under these circumstances was lower. We believe this is To demonstrate the generalizability of our sensing plat- due to the presence of interferant proteins that can non-spe- form, we fabricated sensors employing different recognition cifically adsorb on the surface,[29] thus affecting the collisional elements. Specifically, we replaced the DIG molecules at the activity of the receptors, but not its recognition properties. ends of the recognition strand with two dinitrophenols (DNP) Last, we further demonstrated the programmability of the for the detection of the anti-DNP antibodies (Figure 5A–C). platform by expanding the set of biomolecular targets that can For this sensor, we observed a maximum signal change of be detected with a single-step measurement. We thus selected −71.4 ± 2.1% and a KD of 4.2 ± 0.5 × 10−9 m, thus showing a the multivalent protein streptavidin as the target, and we Adv. Funct. Mater. 2022, 32, 2201881 2201881 (6 of 11) © 2022 The Authors. Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 2022, 37, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202201881 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [14/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.afm-journal.de modified the recognition strand with two biotin molecules. As with the previous sensor, the resulting sensor is specific and selective and can detect the selected target at low nanomolar concentrations (Figure 5D,F, purple curve). The streptavidin sensor displays a signal gain of −32.5 ± 1.2% and a KD of 2.1 ± 0.4 × 10−9 m. Conversely, for this protein target, the mod- ular variant displays a higher signal gain and affinity than the native receptor (Figure S6, Supporting Information). This dif- ference can be explained by considering the structure of the streptavidin where the two binding sites on the same face are ≈2.0 nm apart.[61,62] This distance is very small and only the flex- ibility of the modular variant can support a bivalent binding with the streptavidin. The sensor proved to be specific and to work perfectly with untreated biological fluids (Figure 5E,F). When it was challenged in artificial saliva, we observed a maximum signal change of −32.5 ± 1.2% and a KD of 2.0 ± 0.3 × 10−9 m. On the other hand, in serum it produced a maximum signal change of −19.1 ± 1.3% and a KD of 1.3 ± 0.4 × 10−9 m. These results confirm the ability of the platform to detect different molecular targets with the same efficiency and accuracy. 3. Conclusions Exploiting the programmability of DNA as a nanomaterial and the multivalent binding properties of natural antibodies, we have developed a generalizable, programmable, and versatile electrochemical sensing platform for the single-step detection of biomolecular targets. We designed a Y-shaped DNA “scaf- fold” nanostructure that mimics the shape and the conforma- tional variability of IgG bivalent antibodies. To adapt it to support electrochemical readouts and to detect the selected target, we modified the nanostructure with a thiol group and redox tag molecules, and two specific recognition elements that can be changed in relation to the selected target. The binding of the target antibody or protein with the recognition molecules leads to the formation of a bulky target-receptor complex that reduces the collisional dynamics of the Y-shaped DNA nanostructure, Figure 5. New targets can be detected by changing the recognition ele- producing a decrease in the faradic current that is related to the ments located in the arms of the scaffold receptor. A) We character- concentration of the target. This novel DNA scaffold sensor ena- ized the specificity of a Y-shaped DNA-scaffold sensor presenting two bles, in principle, the detection of any biomolecular target for dinitrophenol (DNP) molecules as recognition elements. The results which two recognition molecules can be incorporated into the obtained confirmed the specific response only in the presence of the sequence of the anchoring strand. To prove this, we used our selected anti-DNP antibodies. B) The sensor responded promptly to the antibody target when challenged in undiluted serum (orange line) sensing platform to quantitively detect two bivalent antibody tar- and artificial saliva (blue line). C) Representative square wave voltam- gets and one multivalent protein target. The sensor can rapidly mograms collected in the absence (gray) and in the presence of the (in <15 min) detect such targets at low nanomolar concentrations antibody target (colored), for the different matrices (for the detection with a high specificity. Lastly, we demonstrated the robustness of of 100 × 10−9 m anti-digoxigenin antibody in buffer, serum and artificial our Y-shaped scaffold sensor by detecting selected biomolecular saliva). D) Similarly, we fabricated a Y-shaped DNA-scaffold sensor pre- targets in complex biological fluids, such as serum and artificial senting two biotin molecules to program its binding with the multivalent protein streptavidin. Again, the sensor only responded to the presence saliva. Despite the sensor can detect the targets in a single step of the protein target. E) The platform enables the single-step detection in these biological fluids, artificial saliva does not have the same of the selected target, and also worked in undiluted serum (orange line) complexity as a pooled human saliva sample. This represents a and artificial saliva (blue line). Both sensors therefore respond to their limitation of this study; therefore, further tests need to be per- specific target at low nanomolar concentrations with a high specificity formed to characterize and optimize the sensor in real saliva. and accuracy. The error bars reflect standard deviations derived using at The development of the electrochemical Y-shaped scaffold least three independently fabricated sensors. F) Representative square sensor offers a new analytical tool compared to the existing wave voltammograms collected in the absence (gray) and in the pres- ence of the proteiin target (colored), for the different matrices (for the methods for detecting biomolecules. For example, the lack of detection of 100 × 10−9 m anti-digoxigenin antibody in buffer, serum and an enzymatic amplification step makes our platform less sen- artificial saliva). sitive than ELISA[8] or other sensing approaches based on the Adv. Funct. Mater. 2022, 32, 2201881 2201881 (7 of 11) © 2022 The Authors. Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 2022, 37, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202201881 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [14/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.afm-journal.de co-localization or binding-induced conformational changes two molecular receptors (whether Epitope of Digoxigenin, Dinitrophenol that can trigger and modulate enzymatic activity.[54,63,64] Despite or Biotin) at both 5′ and 3′ end. Modular and modular internal variants this, the signal transduction mechanism of our sensor makes had the same two recognition strand sequences (recognition strand left and recognition strand right, which were modified with epitopes or the detection of the target reagentless, rapid, and single-step biotin in 3′ or 5′ end, respectively). that is crucial for high-frequency testing. In fact, to perform The sequences and modification schemes are as follows: the measurement, there is no need for washing or incubation, Anchoring strand 1 or the addition of reagents, and there is no risk of environ- 5′-(OH-C6-SS)-GTCAACCTATCGTACTTGCGTTTTTGCTATGTCAACTGCG mental factors affecting the enzymatic catalysis. The analytical TCTAC-(Atto-MB2)-3′ performance of our platform compares well with previously engineered DNA-based electrochemical sensors[24,27,33,54] and Anchoring strand 2 5 ′ - G C A C T C T G ATAT G A C A G A C G T T T T T C G C A A G TA C G ATA G G T more recently developed optical homogeneous assays.[55,63,65–68] TGAC-(Atto-MB2)-3′ Our electrochemical Y-shaped scaffold sensor could be further improved to support mass testing, for example, using calibra- Recognition strand (DIG) tion-free and dual reporter approaches (i.e., allowing the direct 5′-(DIG)-GTAGACGCAGTTGACATAGCTTTTTCGTCTGTCATATCAGA GTGC-(DIG)-3′ quantification of the target).[69–72] It could also be easily inte- grated with mobile phones or portable electrochemical setups, Recognition strand (DNP) thus, offering a more efficient tool for real-time epidemiology.[69] 5′-(DNP)-GTAGACGCAGTTGACATAGCTTTTTCGTCTGTCATATCAGA GTGC-(DNP)-3′ Recognition strand (Biotin) 5′-(Biotin)-GTAGACGCAGTTGACATAGCTTTTTCGTCTGTCATATCAGA 4. Experimental Section GTGC-(Biotin)-3′ Chemical Reagents and Materials: Reagent-grade chemicals, including Recognition strand (control) phosphate buffered saline tablets (PBS), sodium hydroxide (NaOH), 5′-GTAGACGCAGTTGACATAGCTTTTTCGTCTGTCATATCA­GAGTGC-3′ hydrochloric acid (HCl), 6-mercapto-1-hexanol, sodium chloride (NaCl), ethanol (CH3CH2OH), magnesium chloride (MgCl2), tris(2-carboxyethyl)- Recognition strand Left (DIG), Modular variant. phosphine hydrochloride (TCEP), tris[hydroxymethyl]-aminomethane 5′-GTAGACGCAGTTGACATAGC-(DIG)-3′ hydrochloride (Trizma), ethylenediaminetetraacetic acid (EDTA), and 100% Human Serum (H4522-100ML) were obtained from Sigma-Aldrich Recognition strand Right (DIG), Modular variant. (St. Louis, MO, USA), and were used as received. Sulfuric acid 96% 5′-CGTCTGTCATATCAGAGTGC-(DIG)-3′ (H2SO4) was obtained from Panreac Química S.L.U. (Barcelona, Spain). Recognition strand Left (DNP), Modular variant. Antibody targets such as digoxigenin (DIG) sheep polyclonal antibody 5′-GTAGACGCAGTTGACATAGC-(DNP)-3′ (500 μL of 5 mg mL−1 solution; ab64509) were purchased from Abcam (Germany), while 2,4-dinitrophenol (DNP) goat polyclonal antibody Recognition strand Right (DNP), Modular variant. (1 mL of 1 mg mL−1 solution; 17 160 209) was purchased by Fischer 5′-CGTCTGTCATATCAGAGTGC-(DNP)-3′ Scientific (Madrid, Spain). The antibodies were stored at −20 and 4 °C Recognition strand Left (Biotin), Modular variant. respectively and diluted in PBS (NaCl 137 × 10−3 m, KCl 2.7 × 10−3 m, Na2HPO4 10 × 10−3 m, KH2PO4 1.8 × 10−3 m at pH 7.4) before their use. 5′-GTAGACGCAGTTGACATAGC-(Biotin)-3′ The anti-DIG antibody Fab fragment (1 mg) was obtained from Sigma- Recognition strand Right (DNP), Modular variant. Aldrich (St. Louis, MO, USA) in lyophilized form. It was dissolved in PBS to reach a concentration of 1 mg mL−1 and stored at 4 °C. The 5′-CGTCTGTCATATCAGAGTGC-(Biotin)-3′ Streptavidin (10 mg, 13.3 kDa for each subunit) protein target was Recognition strand Left (DIG), Modular internal variant. purchased from Sigma-Aldrich (St. Louis, MO, USA) and was diluted in milliQ water to reach the final concentration of 145 × 10−6 m and stored 5′-(DIG)-GTAGACGCAGTTGACATAGC-3′ at −20 °C. Artificial Saliva for Medical and Dental Research (Part number Recognition strand Right (DIG), Modular internal variant. 1700-0305) was purchased from LCTech GmbH (Obertaufkirchen, 5′-(DIG)-CGTCTGTCATATCAGAGTGC-3′ Germany) and was used as received. The Ag|AgCl reference electrodes (CHI111/112 reference electrode) and platinum wire (CHI115 counter The sequences in bold represent the poly(T) portion. electrode) were purchased from CH Instruments (Austin, TX, USA). Oligonucleotides: High-performance liquid chromatography purified NUPACK Simulations: NUPACK was used (http://www.nupack.org/) oligonucleotides were purchased from Biomers GmbH (Ulm, Germany) to predict the folding of the DNA single strands Anchoring strand 1, and IBA Lifesciences (Göttingen, Germany) and were received in Anchoring strand 2, and the recognition strand (Figure S1, Supporting lyophilized form. The oligonucleotides were dissolved in TE buffer Information).[47] First, the oligo sequences were analyzed using the (10 × 10−3 m Tris buffer, 1 × 10−3 m EDTA, pH 7.8) at a concentration NUPACK to predict their native folding and the absence of stable of 100 × 10−6 m and then aliquoted and stored at −20 °C. The final secondary structure. Then the sequences to confirm the correct concentration of the oligonucleotides was verified using a Shimadzu assembly of the double stranded “stem” portions and the Y-shaped mod. UV-1900 Spectrophotometer (Duisbuirg, Germany) and a TrayCell nanostructure were progressively combined. To perform the analysis (optical path 0.02 cm) measuring the relative absorbance at 260 nm. following parameters were used: a) temperature: 25 °C; b) number All variants of the Y-shape DNA scaffold share the same sequences and of strand species: 1, 2, and 3; c) maximum complex size: 4; d) oligo were composed of the same Anchoring strand 1, Anchoring strand 2, concentration = 500 × 10−9 m; in advanced options; e) [Na+] = 0.15 m, while the recognition strand was the programmable part (Figure S1A, [Mg++] = 0 m; f) dangle treatment: some. Supporting Information). Specifically, Anchoring strand 1 was modified Gold Wire Electrode Fabrication and Electrochemical Cleaning: To with a thiol-C6-SS group at its 5′ end, and a methylene blue molecule at perform the in-vitro characterization of our Y-shaped scaffold sensors 3′ end, while Anchoring strand 2 was modified at 3′ end with a methylene (Figures 3, 4, and 5; Figures S2, S3, S5, and S6, Supporting Information), blue molecule. The control and native variants (Figure 3A) had the same these were fabricated following an established approach.[19,73] Briefly, recognition strand sequence, which was not modified or modified with segments of bare gold wire with a diameter of 200 µm (99.9% Metals Adv. Funct. Mater. 2022, 32, 2201881 2201881 (8 of 11) © 2022 The Authors. Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 2022, 37, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202201881 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [14/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.afm-journal.de basis – Alfa Aesar, Ward Hill, MA, U.S.A.) were cut (5.5 cm in length), saturating target; Figure S2, Supporting Information) was determined. For and the insulated body of the wires was coated using at least two this, at least three sensors in the absence and presence of a saturating layers of heat-shrink polytetrafluoroethylene insulation tubing (PTFE, concentration of the anti-DIG antibody (100 × 10−9 m) were interrogated. HS Sub-Lite-Wall, 0.02 ± 0.001 in, black opaque). To connect with the The peak currents estimated in the absence of the target correspond to potentiostat, a gold pin (item CHI_PIN; CH Instruments, Austin,  IcVariant values, while the peak currents estimated in the presence of the TX, USA) was soldered to one end of the electrode, and this contact [Target] . These raw current values have been converted target correspond to IcVariant was further coated with insulating connector paint (MG Chemicals, to relative signal change IcVariant %signal using the approach described below Burlington, ON, Canada). To perform the functionalization of the andin Equation 2, and it had been performed for each SWV frequency electrode, the uninsulated end of the gold wire to a final length of 3.5 mm tested.[57,74,75] In addition, to understand the effect of the density of the was cut.Then, An electrochemical cleaning of this uninsulated portion DNA scaffold receptor on the signal gain and the sensitivity of the sensors, using the following two-step protocol was performed: 1) 1000–1500 cycles their electrode surface was functionalized with different concentrations of between −1 and −1.6 V in a solution of 0.5 m NaOH at 1 V s−1 to remove the native variant receptor (25, 75, 150, and 300 × 10−9 m). any residual thiol/organic contaminants on the electrode surface and Binding Curve: Experimental binding curves were performed in 9 mL 2) pulsed between 0 and 2 V for at least 16 000 cycles with a pulse length of working buffer (537 × 10−3 m NaCl, 2.7 × 10−3 m KCl, 10 × 10−3 m of 20 ms (no waiting between pulses) in 0.5 m H2SO4 to increase the Na2HPO4, 1.8 × 10−3 m KH2PO4, at pH 7.4) (Figures 3C, 4A, and 5A,C; electrode roughness, as previously reported.[19,73] Figures S3, S5, and S6, Supporting Information) or in 9 mL artificial Electrode Functionalization: The cleaned gold wire electrodes were saliva (Figures 4C and 5B,D), or in 9 mL of human serum (Figures 4C functionalized with the assembled DNA scaffold nanostructure. and 5B,D). For each titration test, at least three gold wire electrodes Specifically, for the Native variant (Figures 2, 3, 4, and 5; Figures S2, S3, modified with the native variant (Figures 3C, 4, and 5; Figures S3, S5, S5, and S6, Supporting Information) and Control Variant (Figure 3C), and S6, Supporting Information) or control variant (Figure 3C), or the following protocol were used: 1) The recognition strand and with the modular variant (Figures 3C; Figures S5 and S6, Supporting Anchoring strand 2 were diluted separately with the assembling buffer Information) or internal modular variant (Figure 3C) were used. More (10 × 10−3 m sodium phosphate, 1 m NaCl, 1 × 10−3 m MgCl2, pH 7.3) specifically, in the absence of the protein target or interferents, a to reach a concentration of 3 × 10−6 m. For their assembly, these preliminary treatment was performed in the working buffer (or in the were mixed using the same volume to obtain a final concentration selected biological fluids) by interrogating the sensors with 30–60 scans of 1.5 × 10−6 m. This solution was heated to 90 °C for 5 min and then until stable peak currents were obtained. Once the scaffold sensor’s cooled to room temperature for 1 h. 2) At the same time, the Anchoring signal was stable, increasing concentrations of the selected antibody strand 1 (100 × 10−6 m) was treated with 10 × 10−3 m tris(2-carboxyethyl)- or protein were added to the working solution at 20 min intervals and phosphine hydrochloride (TCEP; dissolved in milliQ water) for 1 h the sensors were interrogated. The SWV signals were collected using at room temperature in the dark. 3) After this incubation step, the a frequency of 90 Hz except for the binding curves obtained in serum Anchoring strand 1 was diluted with the assembling buffer to reach a and artificial saliva (Figure 4C,D) where a frequency of 5 Hz and 30 Hz, final concentration of 3 × 10−6 m. Next, an aliquot of this solution was respectively was used. The peak current of the scaffold sensor at each added to the previous solution containing the Recognition strand and concentration of target was extracted using a recently published script Anchoring strand 2, to obtain a solution with the three modified oligos written in Python that analyzes the voltametric data by subtracting the [Target]−MAx [74] at a 1:1:1 molar ratio and a concentration of 1 × 10−6 m. Finally, the baseline current from the peak maxima (IcVariant ). The resulting authors waited 20 min for them to assemble. data were fitted using a Langmuir equation (single-site binding)[76] in For the Modular (Figure 3; Figures S5 and S6, Supporting Information) Kaleidagraph (Synergy Software). and Modular internal variant (Figure 3C) the following protocol were used: 1) the Anchoring strand 2 and Recognition strand right were diluted separately with assembling buffer to reach a concentration of 4 × 10−6 m. I[ Target ] = I cVariant +  (  [ target ] I [ Target ]−MAx − I  cVariant cVariant ) (1) Next, they were mixed to obtain a final concentration of 2 × 10−6 m. This cVariant   [ target ] + K D   solution was heated to 90 °C for 5 min and then allowed to cool to room temperature for 1 h. 2) At the same time, the Anchoring strand 1 (2 µL) where [target] is the concentration of the antibody or protein target, oligonucleotide was treated with 4 µL of 10 × 10−3 m TCEP in Tris solution [Target] IcVariant is the raw signal current in the presence of protein target, for 1 h at room temperature and in the dark. 3) After this incubation step,  IcVariant the Anchoring strand 1 was diluted with assembling buffer until it reached is the background raw current seen in the absence of the target, a concentration of 4 × 10−6 m, and the Recognition strand left solution [Target]−MAx IcVariant is the raw current signal change seen at saturating target (4 × 10−6 m) was added to obtain a final solution (Anchoring strand concentrations, and KDis the dissociation constant of the surface-bound 1 + Recognition strand left) with a concertation of 2 × 10−6 m. The authors DNA scaffold receptor. waited 5 min to let them assemble. 4) Finally, the two solutions were mixed  Using the IcVariant values estimated from the fit, the raw signalling to achieve a final solution containing the four modified oligos at a 1:1:1:1 %signal current of each sensor into the relative signal change IcVariant using the molar ratio and a concentration of 1 × 10−6 m. Finally, the authors waited following equation was converted: 20 min for them to assemble. The electrochemically cleaned gold wire electrodes were then immersed in 200 µL of these solutions (containing the assembled Y-shaped receptor) for 1 h in the dark. The electrode %signal  I [ Target ] − I   surface was then rinsed with distilled water and incubated overnight IcVariant =  cVariant cVariant  × 100 (2)  IcVariant  at 4 °C in assembling buffer containing 5 × 10−3 m 6-mercaptohexanol, followed by a further rinse with distilled water before use. Sensor Equilibration Time: The scaffold sensor’s equilibration time Electrochemical Experiments: Electrochemical measurements were was determined (Figure 4B) using the above experimental approach performed in a standard three-electrode cell setup containing an Ag/ and interrogating the sensor every 5 s at 90 Hz in a working buffer AgCl (3 m KCl) reference electrode and platinum counter electrode, (537 × 10−3 m NaCl, 2.7 × 10−3 m KCl, 10 × 10−3 m Na2HPO4, 1.8 × 10−3 m using a CHI1030C Multiplexer (CH Instruments, Austin, TX, USA) at KH2PO4, at pH 7.4). More specifically, after obtaining a stable current room temperature. Square Wave Voltammetry (SWV) was performed baseline (10 min), three different concentrations of anti-DIG antibody using a potential window of −0.10 to −0.42 V, a potential step of 0.001 V to the solution (1, 10, and 100 × 10−9 m) were added, and then the and 0.025 V amplitude. voltammetric signal was monitored for over 20 min. Using the previously Frequency Map: To characterize the sensing mechanism of our cited Python script,[74] the data were analyzed using Equation 2 and electrochemical platform, the SWV frequency dependence of the scaffold converted the raw current into the relative % signal change. The sensors’ signal gain (relative signal change when challenged with the observed signal change was then fitted to a single exponential decay Adv. Funct. Mater. 2022, 32, 2201881 2201881 (9 of 11) © 2022 The Authors. Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 2022, 37, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202201881 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [14/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.afm-journal.de in KaleidaGraph (Synergy Software) to obtain the equilibration time [1] J. T. Ladner, N. D. Grubaugh, O. G. Pybus, K. G. Andersen, Nat. constant of the sensor. Med. 2019, 25, 206. Native PAGE Experiments: The correct DNA strand assembly was [2] Z. Meng, S. Guo, Y. Zhou, M. Li, M. Wang, B. Ying, Signal Trans- evaluated by gel native PAGE electrophoresis (Figures 2B and 3B; duct. Target. Ther. 2021, 6, 1. Figures S4, Supporting Information) using 20% TBE polyacrylamide [3] A. Poschenrieder, M. Thaler, R. Junker, P. B. Luppa, Anal. Bioanal. gels (Invitrogen, ThermoFisher Scientific, Carlsbad, California, USA). Chem. 2019, 411, 7607. Electrophoresis was performed in 1X TBE buffer at 150 V (Figures 2B [4] C. S. Wood, M. R. Thomas, J. Budd, T. P. Mashamba-Thompson, and 3B) (or 200 V; Figure S4, Supporting Information) for 120 min in K. Herbst, D. Pillay, R. W. Peeling, A. M. Johnson, R. A. McKendry, a Cell SureLock Mini-cell chamber (ThermoFisher Scientific, Waltham, M. M. Stevens, Nat. 2019 5667745 2019, 566, 467. Massachusetts, USA). GeneRuler Ultra Low Range was used as the [5] A. N. M. Kraay, K. N. Nelson, C. Y. Zhao, D. Demory, J. S. Weitz, molecular weight marker (ThermoScientific, Waltham, Massachusetts, B. A. Lopman, Nat. Commun. 2021, 12, 7063. USA). The gel was stained with SYBR Gold (dissolved in 1X TBE buffer) for 20 min (Thermo Fisher Scientific, Waltham, Massachusetts, USA), [6] C. Royo-Cebrecos, D. Vilanova, J. López, V. Arroyo, M. Pons, then scanned using a Biodoc-H Imaging System transilluminator G. Francisco, M. G. Carrasco, J. M. Piqué, S. Sanz, C. Dobaño, (Somatco, Riyadh, Saudi Arabia) in combination with TS Software. A. L. García-Basteiro, Lancet Reg. Heal. – Eur. 2021, 5, 100119. [7] G. Rosati, A. Idili, C. Parolo, C. Fuentes-Chust, E. Calucho, L. Hu, C. D. C. Castro E Silva, L. Rivas, E. P. Nguyen, J. F. Bergua, R. Alvárez-Diduk, J. Muñoz, C. Junot, O. Penon, D. Monferrer, Supporting Information E. Delamarche, A. Merkoçi, ACS Nano 2021, 15, 17137. [8] C. Parolo, A. S. Greenwood, N. E. Ogden, D. Kang, C. Hawes, Supporting Information is available from the Wiley Online Library or G. Ortega, N. Arroyo-Currás, K. W. Plaxco, Microsystems Nanoeng. from the author. 2020, 6, 13. [9] R. M. Lequin, Clin. Chem. 2005, 51, 2415. [10] E. Engvall, Clin. Chem. 2010, 56, 319. Acknowledgements [11] K. J. Land, D. I. Boeras, X. S. Chen, A. R. Ramsay, R. W. Peeling, Nat. Microbiol. 2019, 4, 46. A..I. and A..B. contributed equally to this work. ICN2 was funded by the [12] C. Parolo, A. Sena-Torralba, J. F. Bergua, E. Calucho, C. Fuentes-Chust, CERCA programme, Generalitat de Catalunya. ICN2 was supported by L. Hu, L. Rivas, R. Álvarez-Diduk, E. P. Nguyen, S. Cinti, the Severo Ochoa Centres of Excellence programme and funded by the Spanish Research Agency (AEI, grant no. SEV-2017-0706). The authors D. Quesada-González, A. Merkoçi, Nat. Protoc. 2020, 15, 3788. acknowledge Consejo Superior de Investigaciones Científicas (CSIC) [13] A. Clifford, J. Das, H. Yousefi, A. Mahmud, J. B. Chen, S. O. Kelley, for the project “COVID19-122” granted in the call “Nuevas ayudas J. Am. Chem. Soc. 2021, 143, 5281. extraordinarias a proyectos de investigación en el marco de las medidas [14] S. Ranallo, A. Porchetta, F. Ricci, Anal. Chem. 2019, 91, 44. urgentes extraordinarias para hacer frente al impacto económico y social [15] A. B. Bezerra, A. S. N. Kurian, C. J. Easley, Anal. Chem. 2021, 93, 198. del COVID-19 (Ayudas CSIC-COVID-19)”. A.I. was supported by PROBIST [16] Y. Xiao, A. A. Lubin, A. J. Heeger, K. W. Plaxco, Angew. Chemie – Int. postdoctoral fellowship funded by the European Research Council (Marie Ed. 2005, 44, 5456. Skłodowska-Curie grant agreement no. 754510). C.P. acknowledges [17] R. Y. Lai, K. W. Plaxco, A. J. Heeger, Anal. Chem. 2007, 79, 229. the Marie Skłodowska-Curie Actions Individual Fellowship; this project [18] C. Parolo, A. Idili, G. Ortega, A. Csordas, A. Hsu, Q. Yang, received funding from the European Union’s Horizon 2020 research B. S. Ferguson, J. Wang, K. W. Plaxco, N. N. Arroyo-Currás, Q. Yang, and innovation programme under the Marie Skłodowska-Curie grant B. S. Ferguson, J. Wang, K. W. Plaxco, ACS Sens. 2020, 5, 1877. agreement No. 795635. C.P. (ISGlobal) also acknowledges support from [19] A. Idili, C. Parolo, R. Alvarez-Diduk, A. Merkoçi, ACS Sens. 2021, 6, the Spanish Ministry of Science and Innovation** and State Research 3093. Agency through the “Centro de Excelencia Severo Ochoa 2019–2023” [20] A. Idili, J. Gerson, T. Kippin, K. W. Plaxco, Anal. Chem. 2021, 93, Program (CEX2018-000806-S), and support from the Generalitat de 4023. Catalunya through the CERCA Program. R.A.D. received financial support [21] A. Idili, J. Gerson, C. Parolo, T. Kippin, K. W. K. W. Plaxco, Anal. Bio- from the EU Graphene Flagship Core 3 Project (No. 881603). anal. Chem. 2019, 411, 4629. [22] A. Vallée-Bélisle, K. W. Plaxco, Curr. Opin. Struct. Biol. 2010, 20, 518. [23] N. Arroyo-Currás, P. Dauphin-Ducharme, K. Scida, J. L. Chávez, Conflict of Interest Anal. Methods 2020, 12, 1288. [24] A. Vallée-Bélisle, F. Ricci, T. Uzawa, F. Xia, K. W. Plaxco, A. Valle, The authors declare no conflict of interest. J. Am. Chem. Soc. 2012, 134, 15197. [25] A. Idili, F. Ricci, A. Vallée-Bélisle, Nucleic Acids Res. 2017, 45, 7571. [26] M. Rossetti, S. Ranallo, A. Idili, G. Palleschi, A. Porchetta, F. Ricci, Chem. Sci. 2017, 8, 914. Data Availability Statement [27] K. J. Cash, F. Ricci, K. W. Plaxco, J. Am. Chem. Soc. 2009, 131, 6955. The data that support the findings of this study are available from the [28] R. J. White, H. M. Kallewaard, W. Hsieh, A. S. Patterson, corresponding author upon reasonable request. J. B. Kasehagen, K. J. Cash, T. Uzawa, H. T. Soh, K. W. Plaxco, Anal. Chem. 2012, 84, 1098. [29] N. E. Ogden, M. Kurnik, C. Parolo, K. W. Plaxco, Analyst 2019, 144, 5277. Keywords [30] J. Das, S. Gomis, J. B. Chen, H. Yousefi, S. Ahmed, A. Mahmud, antibody detection, DNA-scaffolds, nature-inspired biosensors, protein W. Zhou, E. H. Sargent, S. O. Kelley, Nat. Chem. 2021, 13, detection, serology, square wave voltammetry 428. [31] H. Yousefi, A. Mahmud, D. Chang, J. Das, S. Gomis, J. B. Chen, Received: February 16, 2022 H. Wang, T. Been, L. Yip, E. Coomes, Z. Li, S. Mubareka, A. Mcgeer, Revised: May 31, 2022 N. Christie, S. Gray-Owen, A. Cochrane, J. M. Rini, E. H. Sargent, Published online: June 22, 2022 S. O. Kelley, J. Am. Chem. Soc. 2021, 134, 1722. Adv. Funct. Mater. 2022, 32, 2201881 2201881 (10 of 11) © 2022 The Authors. Advanced Functional Materials published by Wiley-VCH GmbH 16163028, 2022, 37, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/adfm.202201881 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [14/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.afm-journal.de [32] D. Kang, C. Parolo, S. Sun, N. E. Ogden, F. W. Dahlquist, [55] S. Ranallo, M. Rossetti, K. W. Plaxco, A. Vallée-Bélisle, F. Ricci, K. W. Plaxco, ACS Sens. 2018, 3, 1271. Angew. Chemie – Int. Ed. 2015, 54, 13214. [33] S. S. Mahshid, S. Camiré, F. Ricci, A. Vallée-Bélisle, J. Am. Chem. [56] P. Dauphin-Ducharme, K. W. Plaxco, Anal. Chem. 2016, 88, 11654. Soc. 2015, 137, 15596. [57] R. J. White, K. W. Plaxco, Anal. Chem. 2010, 82, 73. [34] G. Vidarsson, G. Dekkers, T. Rispens, Front. Immunol. 2014, 5, 520. [58] F. Ricci, R. Y. Lai, A. J. Heeger, K. W. Plaxco, J. J. Sumner, Langmuir [35] H. Ma, R. O’Kennedy, Methods Mol. Biol. 2015, 1348, 7. 2007, 23, 6827. [36] N. C. Seeman, H. F. Sleiman, Nat. Rev. Mater. 2018, 3, 17068. [59] A. De La Escosura-Muñiz, C. Parolo, A. Merkoi, Mater. Today 2010, [37] N. C. Seeman, Annu. Rev. Biochem. 2010, 79, 65. 13, 24. [38] D. Scalise, R. Schulman, Annu. Rev. Biomed. Eng. 2019, 21, 469. [60] K. J. Cash, F. Ricci, K. W. Plaxco, Chem. Commun. (Camb). 2009, 41, [39] N. L. Goddard, G. Bonnet, O. Krichevsky, A. Libchaber, Phys. Rev. 6222. Lett. 2000, 85, 2400. [61] W. A. Hendrickson, A. Pähler, J. L. Smith, Y. Satow, E. A. Merritt, [40] D. Mariottini, A. Idili, M. A. D. Nijenhuis, T. F. A. De Greef, F. Ricci, R. P. Phizackerley, Proc. Natl. Acad. Sci. USA 1989, 86, 2190. J. Am. Chem. Soc. 2018, 140, 14725. [62] C. L. Ren, D. Carvajal, K. R. Shull, I. Szleifer, Langmuir 2009, 25, [41] F. Iacovelli, C. Alves, M. Falconi, F. Oteri, C. L. P. De Oliveira, 12283. A. Desideri, Biopolymers 2014, 101, 992. [63] A. Porchetta, R. Ippodrino, B. Marini, A. Caruso, F. Caccuri, F. Ricci, [42] F. Iacovelli, M. Falconi, B. R. Knudsen, A. Desideri, RSC Adv. 2016, J. Am. Chem. Soc. 2018, 140, 947. 6, 35160. [64] R. Arts, I. Den Hartog, S. E. Zijlema, V. Thijssen, S. H. E. Van Der [43] P. Zhang, X. Liu, P. Liu, F. Wang, H. Ariyama, T. Ando, J. Lin, Beelen, M. Merkx, Anal. Chem. 2016, 88, 4525. L. Wang, J. Hu, B. Li, C. Fan, Nat. Commun. 2020, E11. [65] D. Legendre, P. Soumillion, J. Fastrez, Nat. Biotechnol. 1999, 17, 67. [44] S. Sandin, L. G. Öfverstedt, A. C. Wikström, Ö. Wrange, [66] L. Tian, T. Heyduk, Anal. Chem. 2009, 81, 5218. U. Skoglund, Structure 2004, 12, 409. [67] K. Tenda, B. van Gerven, R. Arts, Y. Hiruta, M. Merkx, D. Citterio, [45] J. Jay, B. Bray, Y. Qi, E. Igbinigie, H. Wu, J. Li, G. Ren, Antibodies Angew. Chemie – Int. Ed. 2018, 57, 15369. 2018, 7, 18. [68] K. J. Oh, K. J. Cash, A. A. Lubin, K. W. Plaxco, Chem. Commun. 2007, [46] X. Zhang, L. Zhang, H. Tong, B. Peng, M. J. Rames, S. Zhang, 46, 4869. G. Ren, Sci. Rep. 2015, 5, 9803. [69] G. Rosati, M. Urban, L. Zhao, Q. Yang, C. de Carvalho Castro e Silva, [47] J. N. Zadeh, C. D. Steenberg, J. S. Bois, B. R. Wolfe, M. B. Pierce, S. Bonaldo, C. Parolo, E. P. Nguyen, G. Ortega, P. Fornasiero, A. R. Khan, R. M. Dirks, N. A. Pierce, J. Comput. Chem. 2011, 32, A. Paccagnella, A. Merkoçi, Biosens. Bioelectron. 2022, 196, 113737. 170. [70] A. Idili, C. Parolo, G. Ortega, K. W. Plaxco, ACS Sens. 2019, 4, 3227. [48] P. S. Kwon, S. Ren, S. J. Kwon, M. E. Kizer, L. Kuo, M. Xie, D. Zhu, [71] H. Li, P. Dauphin-Ducharme, G. Ortega, K. W. Plaxco, J. Am. Chem. F. Zhou, F. Zhang, D. Kim, K. Fraser, L. D. Kramer, N. C. Seeman, Soc. 2017, 139, 11207. J. S. Dordick, R. J. Linhardt, J. Chao, X. Wang, Nat. Chem. 2020, 12, [72] Y. Du, B. J. Lim, B. Li, Y. S. Jiang, J. L. Sessler, A. D. Ellington, Anal. 26. Chem. 2014, 86, 8010. [49] C. Tian, C. Zhang, Methods Mol. Biol. 2017, 1500, 11. [73] N. Arroyo-Currás, K. Scida, K. L. Ploense, T. E. Kippin, K. W. Plaxco, [50] A. R. Chandrasekaran, Nat. Rev. Chem. 2021, 5, 225. Anal. Chem. 2017, 89, 12185. [51] D. Shu, Y. Shu, F. Haque, S. Abdelmawla, P. Guo, Nat. Nanotechnol. [74] S. D. Curtis, K. L. Ploense, M. Kurnik, G. Ortega, C. Parolo, 2011, 6, 658. T. E. Kippin, K. W. Plaxco, N. Arroyo-Currás, Anal. Chem. 2019, 91, [52] Y. Takezawa, S. Yoneda, J. L. H. A. Duprey, T. Nakama, M. Shionoya, 12321. Chem. Sci. 2016, 7, 3006. [75] A. Idili, A. Amodio, M. Vidonis, J. Feinberg-Somerson, [53] M. Galanti, D. Fanelli, F. Piazza, Sci. Rep. 2016, 6, 18976. M. Castronovo, F. Ricci, Anal. Chem. 2014, 86, 9013. [54] M. Rossetti, S. Brannetti, M. Mocenigo, B. Marini, R. Ippodrino, [76] B. E. Fernández de Ávila, H. M. Watkins, J. M. Pingarrón, A. Porchetta, Angew. Chemie – Int. Ed. 2020, 59, 14973. K. W. Plaxco, G. Palleschi, F. Ricci, Anal. Chem. 2013, 85, 6593. Adv. Funct. Mater. 2022, 32, 2201881 2201881 (11 of 11) © 2022 The Authors. Advanced Functional Materials published by Wiley-VCH GmbH