Electrical Actuation of a DNA Origami Nanolever on an Electrode

Nov 7, 2017 - sorry, we can't preview this file. figshare. Share Download. The authors declare the following competing financial interest(s): F.K., W...
0 downloads 13 Views 843KB Size
Communication Cite This: J. Am. Chem. Soc. 2017, 139, 16510-16513

pubs.acs.org/JACS

Electrical Actuation of a DNA Origami Nanolever on an Electrode Felix Kroener,†,‡ Andreas Heerwig,§ Wolfgang Kaiser,‡ Michael Mertig,*,†,§ and Ulrich Rant*,‡ †

Technische Universitaet Dresden, 01069 Dresden, Germany Dynamic Biosensors GmbH, 82152 Planegg, Germany § Kurt-Schwabe-Institut fuer Mess- und Sensortechnik Meinsberg e.V., 04736 Waldheim, Germany ‡

S Supporting Information *

electrode via DNA hybridization, the rod features a 48 base single-stranded overhang at one end. This anchor sequence is complementary to single-stranded capture oligodeoxynucleotides (ODN), which had been preimmobilized via sulfur−gold chemistry on the electrodes of a switchSENSE biochip. Approximately 105−106 ODNs were deposited on an electrode area of 0.01 mm2. The biochip also comprised an integrated flow channel to allow for exchange of analyte solution. For fluorescence detection, the origami was labeled with three green fluorophores at the surface proximal end. The labeling site at the upper origami end was chosen so orientation of the rod with respect to the surface could be monitored by fluorescence energy transfer. Because the photon emission of fluorophores is quenched gradually in the proximity of a metal surface,13 emission intensity indicates the angle between the origami and the surface. High intensity signifies a “standing” orientation, whereas low intensity signifies a “lying” orientation.14 Immobilization and concurrent electrical orientation switching of origami rods is shown in a real-time fluorescence measurement in Figure 1C. Measurements were performed with a switchSENSE DRX2 biosensor instrument for the analysis of electro-switchable DNA layers (cf. Figure S5). Throughout measurement, a low-frequency square-wave alternating voltage was applied to the gold electrodes (±0.2 V vs ITO counter/ reference electrode, f = 0.2 Hz). The baseline recording before origami injection shows a stable background signal. The injection of 2 nM origami at t = 0.5 min results in a signal increase due to background fluorescence from solute dye-labeled origami now present in the flow channel (Figure 1D). Concurrently, an intensity modulation in the fluorescence signal emerges, which is synchronous with the applied alternating voltage (Figure 1D,E). During negative electrode potentials the fluorescence is high, whereas during positive potentials the fluorescence is low. This behavior can be rationalized by (i) directed immobilization, i.e., end-tethering, of origami rods through their anchors, and (ii) electrically induced switching of origami orientations on the surface. At negative potentials, negatively charged origamis are being repelled and stand on the surface; at positive potentials, they are attracted and lie on the surface. The quasi-linear increase in fluorescence modulation amplitude during the immobilization phase reflects mass transport limited association kinetics (Figure 1C). This is expected, because the flow rate (1.5 μL/min) and the origami concentration (2 nM) were chosen low during the immobiliza-

ABSTRACT: Development of electrically powered DNA origami nanomachines requires effective means to actuate moving origami parts by externally applied electric fields. We demonstrate how origami nanolevers on an electrode can be manipulated (switched) at high frequency by alternating voltages. Orientation switching is long-time stable and can be induced by applying low voltages of 200 mV. The mechanical response time of a 100 nm long origami lever to an applied voltage step is less than 100 μs, allowing dynamic control of the induced motion. Moreover, through voltage assisted capture, origamis can be immobilized from folding solution without purification, even in the presence of excess staple strands. The results establish a way for interfacing and controlling DNA origamis with standard electronics, and enable their use as moving parts in electro-mechanical nanodevices.

T

he DNA origami nanostructures have become versatile building blocks in nanotechnology.1−5 To expand their application and use them as moving parts in nanomachines, it is desirable to establish means to actuate them by external stimuli.6−11 Given their high intrinsic negative charge, DNA origamis lend themselves for the manipulation by electric fields; however, although integration of origami components as functional elements in electro-mechanical devices bears an immense potential for novel applications, electrical actuation of origamis on surfaces has not been established. We demonstrate persistent orientation switching of a model origami structure, namely a rod-like nanolever, which is endtethered to a metal electrode and actuated by externally applied voltages. We analyze electrical switching behavior in terms of the origami’s voltage response and resolve the molecular dynamics of its rotation on the microsecond scale. Moreover, immobilization of origamis can be facilitated in a selective manner by applying positive potentials to the surface, which makes it possible to attach origamis in the presence of an abundance of staple strands from crude reaction solution. Figure 1A shows the assay setup of an end-tethered origami rod on a gold electrode. The used origami is a 100 nm long cylinder consisting of a six-helix DNA bundle with a diameter of 6 nm, folded from a truncated, single-stranded M13mp18 virus DNA scaffold (see Supporting Information).12 Correct folding and mechanical rigidity of the origami rod were confirmed by TEM (Figure 1B, Figure S3) and AFM (Figure S4). To allow immobilization of the origami rod onto the surface of a gold © 2017 American Chemical Society

Received: October 11, 2017 Published: November 7, 2017 16510

DOI: 10.1021/jacs.7b10862 J. Am. Chem. Soc. 2017, 139, 16510−16513

Communication

Journal of the American Chemical Society

comparison, we measured rise time of a 16 nm long DNA helix with 48 base pairs, which is 5.7 μs. To put this into perspective, two counteracting factors influencing the origami dynamics must be considered: on one hand, the origami features a significantly higher rotational friction coefficient than the 48bp helix due to its larger dimensions, and consequently, must rotate slower. Hydrodynamic calculations treating the molecules as cylinders rotating around one end19,20 yield a factor of f origami /f 48bp rot rot = 950. On the other hand, the higher negative charge of the origami results in a stronger electric repulsion from the surface (1766bp/48bp = 37), which should accelerate orientation switching. The experiments show 48bp helix is 13 times faster than the origami (5.7 μs/95 μs), which is less than the factor of 26 estimated from the ratio of the accelerating and decelerating effects (950/37). This deviation is expected because the capacitive charging time of the electrode (∼2 μs) represents a lower limit for the response time of the attached molecules. Movement of the molecules must be preceded by buildup of the ionic screening layer (double layer charging), which generates the necessary high field strengths.14,21 Though motion of the 48bp helix is significantly influenced (slowed) by the finite capacitive charging time, the electrode charging is practically instantaneous compared to the origami rise time, and thus the origami speed is only determined by the origami’s intrinsic friction. Voltage response curves (Figure 2B) of the origami rod and the 48bp helix reveal the origami can be manipulated by the applied electric field; in fact, more efficiently than the 48bp DNA helix. The slow voltage sweep probes the steady state response of the molecules to repulsive and attractive fields and indicates the average rod orientation that adjusts for a given applied voltage. Both molecules show the anticipated behavior in that they are repelled from the electrode at negative potentials, and they are attracted to the electrode at positive potentials. However, voltages at which the helix and the origami rod switch their orientation, as well as the steepness of the transitions, are markedly different. First, the origami requires less positive potentials to lie down on the surface, and second, it reacts more abruptly to a minor change in voltage compared to the shorter and less charged 48bp helix. This result can be interpreted considering a theory based on statistical mechanics,16 which treats the DNA as end-tethered charged rods that are subjected to (i) electric interactions with the charged surface, and (ii) entropic effects, which randomize rod orientations. For positive bias, electric interactions attract the negatively charged molecules, but entropy counteracts the adoption of an ordered state in which all molecules are lying. Basically, entropy (Brownian agitation) repels molecules from the surface. To explain the differences between the origami rod and the 48bp helix observed, it is essential to appreciate that, although the magnitude of Brownian agitation is the same for both molecules (assuming that they are one rigid element and have no internal degrees of freedom), the relative importance of entropy is greater for the smaller 48bp helix, because of its weaker electric interactions with the surface. Because the origami carries more charge, it can be manipulated by the applied electric field more efficiently and responds abruptly when the voltage is being swept across the electrode’s effective “potential of zero charge”, i.e., the midpoint of the transition (0.08 V for the origami curve in Figure 2B). Next, we address the question how the origami can be immobilized on the surface. In particular, we will show the origami can be captured from the folding solution, without

Figure 1. Immobilization and electrically induced orientation switching of DNA origami rods on a gold electrode. (A) Schematic of dye-labeled origami rod bound to electrode surface via a 48bp DNA anchor. (B) TEM image of a negatively stained origami rod. (C) Fluorescence intensity recorded from surface during immobilization of origami rods onto electrode. Alternating potential of ±0.2 V was applied vs an ITO counter electrode at 0.2 Hz throughout measurement. Electrically induced orientation switching can be seen in magnified views in panels D, E, F. Origamis had been purified by gel electrophoresis before measurement.

tion step to allow for a straightforward monitoring of the process. After sufficient accumulation of origami on the electrode (app. 30% of the immobilized ODNs, as estimated from fluorescence intensities before and after hybridization), the origami solution in the channel was removed by flowing buffer at a high rate of 200 μL/min across the surface (t = 17 min). A small air separation at t = 18 min served to separate two buffer volumes in the tubing, facilitated complete removal of the origami solution, and caused the switching behavior of the origami layer to stabilize. It is evident from Figure 1C,F origami rods can be switched persistently and with high efficiency on the surface. Molecular dynamics of the orientation switching process were resolved using time-correlated single photon counting. In brief, a 200 Hz square-wave voltage applied to the electrode drives the origami switching at high frequency. Simultaneously, histograms of photon arrival times within each voltage cycle are recorded and integrated over many cycles. Figure 2A shows the upward motion of a layer of origami rods, which are repelled from the electrode when a negative voltage step is applied. The observed motion is overdamped, meaning rotation of the rods is dominated by their hydrodynamic friction. This agrees with findings for shorter DNA helices.15−18 The rise time (defined by a signal change from 10% to 90%) of the 100 nm long origami rod is 95 μs. For

Figure 2. Time resolved (A) and voltage resolved (B) origami orientation switching. Blue: 100 nm long six-helix bundle; black: 16 nm long 48bp DNA helix. (A) Upward orientation switching upon a voltage step from −0.2 to +0.2 V (vs ITO) at t = 0. f = 1 kHz for 48bp, f = 200 Hz for origami, signals were integrated for 1 s. (B) Voltage response, potentials are swept slowly at 25 mV/3 s. n = 4 curves are shown for each measurement. 16511

DOI: 10.1021/jacs.7b10862 J. Am. Chem. Soc. 2017, 139, 16510−16513

Communication

Journal of the American Chemical Society

rate) as was observed for immobilization from purified solution. The increase in the green fluorescence signal in Figure 3A shows adsorption of the green-labeled origami on the surface. Conversely, no signal change is observed when a negative control solution containing only staple strands (no origamis) is injected. The red fluorescence signal shown in Figure 3B shows a concomitant increase with the green signal upon injection of unpurified origami solution. This is in accord with the expectation: as the origami anchor hybridizes to the capture DNA, the capture DNA transforms from a single- to a doublestrand. Double-stranded oligos exhibit greater fluorescence modulation amplitudes than single-strands, because they are more rigid and can be oriented more effectively by the applied electric field.21 Understandably, the magnitude of the signal increase is more pronounced for the green fluorescence originating from the top of the origami, than for the red-labeled capture DNA, which is in closer proximity to the fluorescencequenching metal surface. A noteworthy result of Figure 3 is the flat line of the “staples only” solution in panel B: the staples, especially the free anchor staples, do not bind to the capture DNA under the same conditions where the origami can be captured efficiently. Evidently, for the applied voltage the origami is being immobilized exclusively. To elucidate the effect of applied potential on the selectivity of the origami immobilization, we performed association measurements at different positive potentials, Figure 4. For potentials

purifying the origami from excess staple strands. This is an essential advantage, because standard purification procedures like gel electrophoresis are time-consuming and usually entail a significant loss of origami. However, excess staple strands, which are not used up during the folding by the scaffold strand, can be highly problematic, especially when they serve as anchors for surface attachment of the origami. They overwhelmingly compete with the origami for binding to capture DNA on the surface, because they are present at a higher concentration, and they diffuse more quickly than larger origami (here, after folding, 8 nM excess staples compete with 2 nM folded origami). Studies show oligonucleotide hybridization can be facilitated when positive voltages are applied to the electrodes where the capture DNA is immobilized.22−27 We surmised that if an applied positive voltage would preferentially act on the origami rather than the staple (anchor) strands, it should be possible to attract and immobilize origamis without the need to remove excess staples. To test this hypothesis, we used the dual-color detection mode of the DRX2 instrument. As before, green dyes were attached to the upper end of the origami rod, but in addition, the surface immobilized 48nt capture strands were labeled with red dyes, cf. Figure 3. The red dye is not only sensitive to the presence of

Figure 3. Selective immobilization of origami in the presence of excess staple strands from unpurified folding solution. (A) Fluorescence of the green dye, used to label the origami. (B) Fluorescence of the red dye, used to label the immobilized capture DNA. Arrows mark injections of solutions containing “origami with staples” (2 nM origami and 8 nM excess staples) or a control solution with “staples only” (8 nM staples). A square-wave alternating voltage of ±0.2 V ( f = 1 Hz) was applied.

Figure 4. Voltage dependence of potential-assisted origami immobilization. Voltage values state the positive peak voltage of the applied squarewave pulse, the peak-to-peak magnitude was kept constant (Vpp = 0.4 V, f = 1 Hz). The fluorescence modulation amplitude of the green origami dye-label is analyzed as an indicator for the amount of surfaceimmobilized origami. (A) Real-time association measurements. (B) End-point analysis of the amount of origami immobilized after 5 min.

origami but also to the hybridization of free staple strands to the capture DNA. To exclusively detect surface immobilized molecules, alternating voltages (±0.2 V, 1 Hz) were applied and the fluorescence modulation amplitude ΔF = Fup − Fdown was analyzed. The method is analogous to “lock-in detection” in that only molecules, which undergo an electrically induced orientation switching on the surface exhibit a modulated fluorescence signal and are detected, whereas unmodulated background signals (e.g., from green-labeled staples in solution) are suppressed. The experiments shown in Figure 3 were performed with unpurified origami folding solution, which contained a mixture of folded origami rods (2 nM) and excess staple strands (8 nM). Notably, among the staple strands were anchor sequences, competing with the folded origami for the capture DNA on the surface, as well as green-labeled staples, used to attach the green dyes to the top origami end. Successful origami immobilization from unpurified folding solution is demonstrated by the dual-color measurement in Figure 3, and it proceeds with similar efficiency (immobilization

below +0.1 V, we did not observe a significant origami immobilization during a contact time of 5 min. Applying more positive potentials, the origami association rate increased sharply, which is expected for an immobilization process driven by electrostatic attraction. It is remarkable, however, the association rate shows a maximum at a moderate positive potential (+0.2 V), and then decreases again for more positive potentials. We attribute this to the competition with free anchor staples. In fact, measurements of the association rates of staples from staples-only solutions showed, as expected, immobilization of anchor sequences within the staple solution is facilitated by positive voltages, too (Figure S6). But compared to the origamis, it takes more positive voltages to notably accelerate the immobilization of staples. Though onset for electrically facilitated origami capture is +0.1 V, the onset for the accelerated capture of staple strands is +0.3 V. This can be rationalized by the large charge difference of the two molecules. To confirm this, we 16512

DOI: 10.1021/jacs.7b10862 J. Am. Chem. Soc. 2017, 139, 16510−16513

Communication

Journal of the American Chemical Society

SENSE instruments used in this study. A.H. and M.M. declare no competing financial interests.

performed voltage assisted hybridization measurements with ODNs of different lengths and found also in this case longer ODNs with more negative charge were attracted to the electrode already at less positive voltages (Figure S7). Additional frequency-dependent measurements indicate lower frequencies favor the voltage-assisted immobilization (Figure S5). The results of Figures 3 and 4 show origamis can be immobilized on the surface of positively biased electrodes, even in the presence of an excess of competing free anchor sequences. This is important for the practicality of the surface attachment, because the origamis can be captured from their folding solution, which avoids purification losses and increases assembly yields. For optimal selectivity, a moderate positive voltage should be chosen. At voltages that are too positive, preferential attraction of the origami versus the less charged staple strands diminishes, and competition for hybridization to the capture strands on the surface shifts in favor of free anchor strands. In conclusion, we have shown surface-anchored origamis can be manipulated by applying voltages of only several hundred millivolts to a gold electrode. Importantly, the applied voltages induced a capacitive polarization of the metal/electrolyte interface, but were low enough to avoid Faradaic currents and electrochemical (redox) reactions, which might be detrimental to the biomolecules. Consequently, the origami was stable and could be switched persistently over many hours. The rotation time of the 100 nm long rod was measured to be ∼100 μs, proving fast response times can be achieved for electrical actuation of moving origami parts. Regarding utility of the actuation scheme for operation of origami nanomachines, it is remarkable three key requirements are fulfilled: First, energy is provided from an inexhaustible external source (low-cost voltage source or frequency generator), enabling continuous operation. Second, the motion can be controlled with minimal delay, i.e., on the time scale of many molecular processes. Third, by physically attaching the origami to an electrode, an interface between the nanoscale component and a tangible macroscopic electrical circuit is established. Chemically fueled molecular machines, in comparison, usually require fluidic systems involving slow mass transport and meet these requirements only in part. We expect these results can be transferred to more sophisticated origami systems and will pave the way for construction of origami machines that can be turned ON and OFF at the flip of a switch.





ACKNOWLEDGMENTS The authors gratefully acknowledge financial support by the Bundesministerium für Wirtschaft und Energie (BMWi/ZIM: ZF4131001MD5 and ZF4088702MD5), the German Excellence Initiative via the DFG Cluster of Excellence EXC “Center for Advancing Electronics Dresden” (cfaed, EXC1056/1), and the ESF project MindNano (100226937).



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b10862. Experimental details (PDF)



REFERENCES

(1) Rothemund, P. W. K. Nature 2006, 440 (7082), 297. (2) Castro, C. E.; Su, H.-J.; Marras, A. E.; Zhou, L.; Johnson, J. Nanoscale 2015, 7 (14), 5913. (3) Kauert, D. J.; Kurth, T.; Liedl, T.; Seidel, R. Nano Lett. 2011, 11 (12), 5558. (4) Mathieu, F.; Liao, S.; Kopatsch, J.; Wang, T.; Mao, C.; Seeman, N. C. Nano Lett. 2005, 5 (4), 661. (5) Henning-Knechtel, A.; Wiens, M.; Lakatos, M.; Heerwig, A.; Ostermaier, F.; Haufe, N.; Mertig, M. Beilstein J. Nanotechnol. 2016, 7 (1), 948. (6) Marras, A. E.; Zhou, L.; Su, H.-J.; Castro, C. E. Proc. Natl. Acad. Sci. U. S. A. 2015, 112 (3), 713. (7) Funke, J. J.; Dietz, H. Nat. Nanotechnol. 2015, 11 (1), 47. (8) Ketterer, P.; Willner, E. M.; Dietz, H. Sci. Adv. 2016, 2 (2), e1501209. (9) Langecker, M.; Arnaut, V.; Martin, T. G.; List, J.; Renner, S.; Mayer, M.; Dietz, H.; Simmel, F. C. Science (Washington, DC, U. S.) 2012, 338 (6109), 932. (10) Plesa, C.; Ananth, A. N.; Linko, V.; Gülcher, C.; Katan, A. J.; Dietz, H.; Dekker, C. ACS Nano 2014, 8 (1), 35. (11) Seifert, A.; Gopfrich, K.; Burns, J. R.; Fertig, N.; Keyser, U. F.; Howorka, S. ACS Nano 2015, 9 (2), 1117. (12) Said, H.; Schüller, V. J.; Eber, F. J.; Wege, C.; Liedl, T.; Richert, C. Nanoscale 2013, 5 (1), 284. (13) Chance, R. R.; Prock, A.; Silbey, R. Advances in Chemical Physics 1978, 37, 1−65. (14) Kaiser, W.; Rant, U. J. Am. Chem. Soc. 2010, 132 (23), 7935. (15) Langer, A.; Hampel, P. A.; Kaiser, W.; Knezevic, J.; Welte, T.; Villa, V.; Maruyama, M.; Svejda, M.; Jahner, S.; Fischer, F.; Strasser, R.; Rant, U. Nat. Commun. 2013, 4, 2099. (16) Langer, A.; Kaiser, W.; Svejda, M.; Schwertler, P.; Rant, U. J. Phys. Chem. B 2014, 118 (2), 597. (17) Sendner, C.; Kim, Y. W.; Rant, U.; Arinaga, K.; Tornow, M.; Netz, R. R. Phys. Status Solidi A 2006, 203 (14), 3476. (18) Rant, U.; Arinaga, K.; Tornow, M.; Kim, Y. W.; Netz, R. R.; Fujita, S.; Yokoyama, N.; Abstreiter, G. Biophys. J. 2006, 90 (10), 3666. (19) Tirado, M. M.; Martínez, C. L.; de la Torre, J. G. J. Chem. Phys. 1984, 81 (4), 2047. (20) Ortega, A.; García de la Torre, J. J. Chem. Phys. 2003, 119 (18), 9914. (21) Rant, U.; Arinaga, K.; Fujita, S.; Yokoyama, N.; Abstreiter, G.; Tornow, M. Nano Lett. 2004, 4 (12), 2441. (22) Edman, C. F.; Raymond, D. E.; Wu, D. J.; Tu, E.; Sosnowski, R. G.; Butler, W. F.; Nerenberg, M.; Heller, M. J. Nucleic Acids Res. 1997, 25 (24), 4907. (23) Sosnowski, R. G.; Tu, E.; Butler, W. F.; O’Connell, J. P.; Heller, M. J. Proc. Natl. Acad. Sci. U. S. A. 1997, 94 (4), 1119. (24) Heaton, R. J.; Peterson, A. W.; Georgiadis, R. M. Proc. Natl. Acad. Sci. U. S. A. 2001, 98 (7), 3701. (25) Gurtner, C.; Tu, E.; Jamshidi, N.; Haigis, R. W.; Onofrey, T. J.; Edman, C. F.; Sosnowski, R.; Wallace, B.; Heller, M. J. Electrophoresis 2002, 23 (10), 1543. (26) Jambrec, D.; Gebala, M.; La Mantia, F.; Schuhmann, W. Angew. Chem., Int. Ed. 2015, 54 (50), 15064. (27) Wong, I. Y.; Melosh, N. A. Nano Lett. 2009, 9 (10), 3521.

AUTHOR INFORMATION

Corresponding Authors

*[email protected] *[email protected] ORCID

Michael Mertig: 0000-0002-8359-0135 Ulrich Rant: 0000-0002-6100-828X Notes

The authors declare the following competing financial interest(s): F.K., W.K., and U.R. are employees of Dynamic Biosensors GmbH, which is the manufacturer of the switch16513

DOI: 10.1021/jacs.7b10862 J. Am. Chem. Soc. 2017, 139, 16510−16513