Programmable DNA Nanosystem for Molecular Interrogation

Abstract

We describe a self-assembling DNA-based nanosystem for interrogating molecular interactions. The nanosystem contains a rigid supporting dumbbell-shaped frame, a cylindrical central core, and a mobile ring that is coaxial with the core. Motion of the ring is influenced by several control elements whose force-generating capability is based on the transition of single-stranded DNA to double-stranded DNA. These forces can be directed to act in opposition to adhesive forces between the ring and the frame thereby providing a mechanism for molecular detection and interrogation at the ring-frame interface. As proof of principle we use this system to evaluate base stacking adhesion and demonstrate detection of a soluble nucleic acid viral genome mimic.

Programmable DNA Nanosystem for Molecular Interrogation

OPEN

Divita Mathur,,, & Eric R. Henderson,

We describe a self-assembling DNA-based nanosystem for interrogating molecular interactions. The nanosystem contains a rigid supporting dumbbell-shaped frame, a cylindrical central core, and a mobile force-generating capability is based on the transition of single-stranded DNA to double-stranded DNA. These forces can be directed to act in opposition to adhesive forces between the ring and the frame thereby providing a mechanism for molecular detection and interrogation at the ring-frame interface. As proof of principle we use this system to evaluate base stacking adhesion and demonstrate detection of a soluble nucleic acid viral genome mimic.

In addition to the iconic genetic code, nucleic acid contains an engineering code. Recent conceptual and methodological advances have culminated in the availability of tools and strategies for leveraging this engineering code to program DNA to spontaneously create a diverse array of two- and three-dimensional shapes15. These shapes can be imbued with information and function including algorithmic calculations68, single-molecule analyses9,10, therapeutics1113, mechanical actuation14,15, and a host of other capabilities1619. Programmable, responsive actuation of dynamic self-assembling nanodevices is a highly desirable attribute and several studies have demonstrated mechanical reconguration of DNA nanodevices by thermal motion14 and upon triggering by a target molecule11,2026. The study described here focuses on improvements in these responsive nanosystems in terms of modularity and robustness while minimizing undesirable conformational changes upon actuation. The DNA nanosystem described here enhances the repertoire of molecular reporting systems25,27 and serves as a base platform for molecular interrogation with an embedded reporter system module that is compatible with a variety of molecular species.

In this study we harness the dierence in persistence length (i.e., rigidity) of single-stranded (ssDNA) and double-stranded (dsDNA) DNA to elicit a dened physical state change in a self-assembling DNA nanosystem we have termed OPTIMuS (Oligo-Propelled Technology for Interrogating Molecular Systems; Fig.1). This inducible state change can be used to interrogate user-programmed molecular interactions within the OPTIMuS platform. In this report, we demonstrate how OPTIMuS can be used to detect a soluble target molecule and assess the relative strength of a non-covalent (base stacking) molecular interaction.

Results

Construction and Principle of Operation. The OPTIMuS platform is constructed following the principles of DNA origami, in which, specic ensembles of short oligonucleotides called staples are used to fold a large single-stranded scaold into desired shapes (for details see Supplementary Information & Supplementary Fig. 2). The mechanical design of OPTIMuS is inspired by a system in which tunable springs exert pushing and pulling forces on a movable ring within a coaxial dumbell-shaped framework (Fig.1a). These forces are opposed by introducing resistance at the interface between the mobile ring and the dumbbell frame. Finally, an embedded Frster Resonance Energy Transfer (FRET) system, in which one cyanine 3 (cy3) molecule is positioned on the frame and one cyanine 5 (cy5) molecule is on the ring, reports the relative position of the ring under various conditions (Supplementary Fig. 9). The main components of OPTIMuS are described in Fig.1b.

The force elements are single-stranded scaold domains that undergo structural change upon hybridizing to their complementary staple strands. Single-stranded DNA, an entropically elastic polymer with a formal contour length of 0.7nm/base and persistence length of about 5nm28, transitions into a rigid double-helix of 0.34nm/bp contour length and 50 nm persistence length upon hybridizing with its complimentary strand29 (Fig.1e). If the

College of

A P

SCIENTIFIC REPORTS

1

www.nature.com/scientificreports/

Figure 1. OPTIMuS operational principles. (a) A schematic illustrating the key mechanical components of OPTIMuS. A movable ring surrounds a cylindrical core that is anchored at both ends by dumbbell shaped frame elements. The ring is pushed and/or pulled away from frameR by user-controlled ssDNA to dsDNA transitions, whereas resistance at the interface can obstruct ring movement. (b) A three-dimensional rendering of OPTIMuS showing the 24 helix bundle in the honeycomb lattice arrangement (Supplementary Figs 2 and 4). (c) A cross-sectional view of the ring/frameR interface shows active sites, loops and the FRET reporter pair. (d) The idealized overall reconguration that can be elicited in OPTIMuS. On the le is the ground state (G) that has all force domains in single-stranded form (EC, Ch, L). Upon adding staples corresponding to them (ECS, ChS, LS) the nanosystem recongures with a displaced ring position. Hybridization on the le-side (atEC and Ch) and the right-side (L) of the ring is reported by via FRET. (e) Depiction of the mechanism of force-induced motion by an ssDNA to dsDNA transition. The three force domains, EC, Ch and L are based uponthe following scenarios. In scenario 1, hybridization of randomly coiled ssDNA creates a pushing force that increases the separation between attached substrates. In Scenario 2, pulling forces are created when a stretched ssDNA collapses into a short double helix upon hybridization, thereby bringing the substrates closer together.

In scenario 3, which occurs when EC is formed before or simultaneously with Ch duplexes, the mechanically stretched ssDNA cannot form a duplex with its complement despite the favorable G for the same molecules when stereochemically unconstrained. (f) A schematic of the blunt end and non-blunt end interactions between the coaxial helices of ring and frameR. Non-blunt ends are created by leaving eight scaold bases at the crossover unhybridized. The resultant single-stranded region prevents base stacking and minimizes adhesive interaction between the duplexes. (g) Illustration of a toehold-mediated DNA strand displacement reaction.

ends of an ssDNA molecule are tethered to two substrates, the relative distance between the substrates can be altered due to the internal reconguration concomitant with duplex formation (Fig.1e Scenario 1 and 2). This spatial change can be exploited for applications in sensing30,31 and, potentially, molecular force/energy measurements. Importantly, although somewhat counterintuitive, if these substrates are immobilized, causing the ssDNA to be xed at its full extenstion limit (0.7nm/base), duplex formation becomes stereochmically inhibited despite the favorable G of the reaction (Fig.1e Scenario 3).

Three kinds of scaold domains in OPTIMuS use the aforementioned phenomena to impart pushing or pulling force on the central ring to move it from a position proximal to the right side of the frame (frameR) toward the le side of the frame (frameL) (Fig.1d). These domains are termed extended core (EC), cinchers (Ch) and loops (L). The corresponding staple strands are termed ECS, ChS, LS, respectively. The L domain function is illustrated in Fig.1e, scenario 1. Upon hybridization to LS staples, the L domain extension causes the ring to move away from frameR. The Ch domain contains stretches of scaold DNA that are shorter in length (35 bases) than the underlying EC domain (70 bases) (Supplementary Fig. 3). As shown in Fig.1e, scenario 2, these domains pull the ring towards frameL. Finally, extension of the EC domain by hybridization to ECS staples generates a pushing force on frameL, thereby moving it away from frameR. This motion results in extension of cincher DNA which

SCIENTIFIC REPORTS

2

www.nature.com/scientificreports/

Figure 2. System characterization and demonstration using FRET and Transmission Electron Microscopy (TEM). (a) A TEM eld showing examples of the ground state of OPTIMuS. White arrows adjacent to some structures indicate single-stranded clouds of DNA corresponding to the unhybridized force domains. (b) FRET output of OPTIMuS as a function of combination of pre-added (during self-assembly) force domain components. FRET is a reliable indicator of ring/frameR distance and, therefore, a reporter of force-induced ring motion.

A schematic of each conguration is shown for clarity. In case of G+ECS, pre-adding ECS aects the stable formation of the structure due to internal tension between Ch and EC, thereby destabilizing the formation of the frame (Supplementary Fig. 1). This is one type of internal tension that comprises the foundation of mechanical actuation in the nanosystem. (c) Corroborating TEM structures corresponding to key OPTIMuS congurations used in this study (Scale bar= 50nm). Green and yellow arrows indicate structural gaps corresponding to ssDNA L and ssDNA Ch domains respectively.

pulls on the ring, moving it away from frameR (Supplementary Fig. 3). These design features are illustrated in caDNAno layouts corresponding to each conguration in Supplementary Figs 47.

Motion of the ring induced by hybridization of force domains can be challenged by introducing resisting adhesive force(s) at the interface (active sites) of the ring and frameR (Fig.1c). Two types of resistance were tested in this study. The rst type is blunt end base stacking (pi-bond interation) at the active sites (Fig.1f). Previous work has shown that base stacking interactions can result in strong helix-helix adhesion and this interaction has been utilized to create multiunit self-assembling DNA nanostructures32,33. Previous studies also suggest that GC dinucleotides form the strongest stacking interaction34. Therefore, the nanosystem design used in this study employed GC dinucleotide base stacking to promote adhesion at the active sites (Supplementary Fig. 10).

Another kind of resistence is a lock and key system, the design of which is based on toehold-mediated strand displacement25 (Fig.1g). Toehold-mediated strand displacement is a dynamic hybridization event wherein a DNA strand invades and displaces another strand from a duplex by binding to a short single-stranded olionucleotide extension called the toehold. In these experiments the active sites were decorated with toehold-bearing duplexes that tether the ring to frameR.

System Characterization and Demonstration. To assess the scope of inuence of each force inducing domain (EC, Ch, L) on the central ring and its eect on the ensemble FRET output, we assembled the nanosystem in the presence of dierent combinations of domain-specic staples (Fig.2). OPTIMuS at ground state (G) is dened as the conguration that contains no force domains (EC, Ch, and L remain single-stranded). In each sample, staples associated with G (i.e., frameR, frameL, core and ring) were mixed with active site staples conferring blunt end formation and also a specic combination of force domain staples. Aer mixing, the nanosystem was assembled following the standard protocol (Methods). The FRET output reects the position of the ring relative to frameR in each conguration (corroborated by Transmission Electron Microscopy (TEM) analysis; Fig.2). Since blunt end stacking is strongly distance-dependent35, and therefore cannot acquire sufficient force when assembled simultaneously with opposing force domains, the force domains prevail in all cases except ChS (see following) and induce various degrees of ring movement. Figure2b shows that ChS alone has no eect on FRET compared to G. In contrast, LS alone and ECS + LS elicit the same FRET as all three force domains in combination (ECS+ChS+LS). Finally, G+ECS and G+ECS+ChS have comparable FRET outputs. Thus, these two force domain combinations, while diering in total G appear to reach a common nal mechanically limited state of the nanosystem (minimum FRET). Based on these results, we chose three combinations of force domain staples

SCIENTIFIC REPORTS

3

www.nature.com/scientificreports/

Figure 3. Base stacking adhesive forces versus OPTIMuS actuators. (a) The interface makeup between the ring and frameR shown as a cross-section of OPTIMuS. It can be modied with coaxial blunt ends (BE) to create high adhesion or no blunt ends to minimize adhesion. (b) FRET output of all blunt ends (All BE) versus no blunt ends (No BE) OPTIMuS upon the addition of dierent force domains (****Indicates P0.0001; *Indicates P0.05).

as actuators, ECS, ECS+ ChS and ECS+ LS, to study their eect on the motion of the ring when opposed by two dierent kinds of resistant forces, base stacking and DNA-DNA hybridization.

Base stacking vs Force domains. Figure3 shows the results of experiments in which combinations of force inducing domains were tested for their ability to disrupt base stacking-mediated ring/frameR adhesion. Ground state with blunt ends on all active sites (called GAll BE) and ground state with no blunt ends on all active sites (called GNo BE) were initially assembled in the absence of force domains. The samples were divided into equal amounts, ltered to remove excess staples, and then incubated with either buer alone, ECS, ECS+ChS or ECS+LS (Methods).

In the case of GAll BE (all blunt ends stacked) the combination of duplex formation in EC and L domains resulted in rupture of the adhesive bond between the ring and frameR. In contrast, ECS + ChS failed to disrupt the base stacking interaction. We hypothesize that this is the consequence of fully-extended and strained cincher ssDNA being unable to form a sufficient number of hydrogen bonds to initiate an ssDNA to dsDNA transition and thereby create a resultant pulling force (in contrast to ease of duplex formation when cinchers are hybridized during the initial self-assembly process, Fig.1e Scenario 3; Fig.2b). Unlike strained cincher domain ssDNA, ssDNA in the loop domain is not stretched but, rather, randomly coiled and, therefore, stereochemically available for hybridization with complementary oligonucleotides. Thus, hybridization to loop domains generates a pushing force on the ring that is sufficient to rupture of the ring/frameR adhesive interaction.

When OPTIMuS was congured to lack base stacking interactions at the ring/frameR interface (Fig.3b GNo BE) all combinations of force domains tested were able to induce ring displacement. In particular, duplex formation of EC+ Ch was able to disrupt the ring/frameR interface because there was no opposing adhesive force and, therefore, no hyperextension of the ssDNA-cincher domains to preclude cincher duplex formation (Fig.1e Scenario 2). The results of these experiments suggest that OPTIMuS may be useful for interrogating other types of molecular interactions at the ring/frameR interface.

TEM analysis was carried out to corroborate FRET analysis of the various configurations of OPTIMuS. Figure4 shows that with full bunt end stacking (GAll BE ) internal reconguration does not take place in the presence of ECS or ECS + ChS (Fig.4a(iiii)). However, the addition of subsequent force domains (ECS + LS) induces a change that leads to ring movement (Fig.4a(iv)). This can be vered by observing the gap inside OPTIMuS which shows the ssDNA cincher domain and helps locate the relative position of the ring (TEM images, Fig.4a(iv)). The overall length of OPTIMuS is also a good indicator of internal reconguration, as can be seen by the dimensional analysis of a population of each kind of sample (right columns, Fig.4a,b). The mean length of GAll BE remains the same upon addition of ECS, but shis to an intermediate length in case of ECS+ChS. The partial hybridization aects the overall length of the nanosystem, but is unable to cause motion in the ring, hence the FRET signal does not alter. This observation supports our hypothesis that there is partial hybridization in the two force domains, EC and Ch, but complete hybridization is stereochemically hindered (i.e., Ch is physically constrained in a stretched conguration) by the blunt end stacking at the active sites.

In contrast to the results above, the conguration lacking blunt end stacking, GNo BE, undergoes an incremental shi in the ring position as well as length of the nanosystem as a function of force domain hybridization (Fig.4b). The addition of ECS alone causes OPTIMuS to assume a bent conguration (Fig.4b(ii)). This can be attributed to a fully-extended ssDNA Ch domain, the tension in which is sufficient to distort the otherwise linear core architecture. This bending serves as evidence that hybridization of the ECS to EC domain is occuring with high efficiency. Duplex formation in the remaining two force domain samples, ECS + ChS, ECS + LS, results in full shi in the rings position and a corresponding overall increase in the length of the nanosystem. The conguration-specic gaps corresponding to single-stranded Ch and L domains permit unambiguous orientation determination of the molecule and indicate the internal position of the ring (Fig.4b(iii,iv)).

SCIENTIFIC REPORTS

4

www.nature.com/scientificreports/

Figure 4. Base Stacking Adhesive Forces versus OPTIMuS actuators: TEM analysis. For each sample, a two-dimensional (2D) rendering was generated using corresponding caDNAno layout, followed by TEM imaging, extraction of three representative images and histrograms showing the density distribution of the overall length of OPTIMuS populations in various congurations. Red curve represents the Normal distribution, blue curve shows the actual distribution and the green dashed line indicates the average length of OPTIMuS in the corresponding pre-added force domain conguration (based on data acquired in Fig.2). (a) TEM images of ground state containing all blunt ends in the presence of dierent force domains. (b) A schematic of the blunt end and non-blunt end interactions between the coaxial helices of ring and frameR.

Detection of a soluble ssDNA viral genome mimic. The earlier suggestion that OPTIMuS might serve as a useful molecular detection platform was tested using a strand displacement mechanism and a viral genome (DNA) mimic. Toehold-bearing duplexes, called ligand domains, were integrated with the ring/frameR interface such that one strand of each duplex extends from the ring and the other from frameR. The toehold-containing strand was designed to be complementary to a soluble target oligonucleotide based on five Ebola genome sequence elements (Supplementary Information). The active sites were modied with these ligand domains (Fig.5a,b). Addition of the target strands disrupted the ligand duplex through toehold-mediated DNA strand displacement25, thereby reducing ring/frameR adhesion and permitting free motion in the ring. We compared ring motion in the presence and absence of target molecules in dierent OPTIMuS congurations. Following the format of experiments that tested blunt end stacking interactions in Figs3 and 4, we constructed the ground state in the absence of the force domains (GNo BE) but bearing the ligand duplexes at the active sites (Fig.5a). The ground state sample was divided, puried via ltration to remove excess staples, and incubated (Methods) with the force domains with or without target strands and FRET was measured. Upon force induction in the absence of target the FRET signal only changed in the presence of the strong disruptive force domain combination ECS+LS.

However, in the presence of soluble target the FRET signal was signicantly reduced when the ring was induced to move using the weaker force-generating domains EC or EC+ Ch, thereby illustrating that the OPTIMuS platform has the potential to serve as a molecular detection platform (Fig.5c).

Discussion

We describe a self-assembling DNA nanosystem termed OPTIMuS that is capable of interrogating molecular interactions by exerting user-controllable forces to challenge the molecular system of interest. Controlled exertion of force in OPTIMuS is founded on the inherent elasticity of ssDNA (a relatively weak and compliant spring), the shortening and stiening of double helical DNA (a relatively strong and sti spring), the high specicity of DNA base pairing, and the adhesive force exhibited by base stacking. The availability of a plurality of control elements should allow OPTIMuS to be tuned to interrogate interactions of a range of strengths. In the present study we explore a soluble target strand displacement mechanism of detection and the interactive force present in base-stacked DNA duplexes.

SCIENTIFIC REPORTS

5

www.nature.com/scientificreports/

Figure 5. Programming OPTIMuS for nucleic acid detection. (a) Cross-section of the interface displaying the placement of the target-associated duplexes. Five sites were remodeled to contain unique duplexes such that the toehold-bearing strand emerged from a staple on the ring and its complement emerged from frameR,

as shown in the 3D rendering. The duplex contributes to ring-frameR adhesion. The remaining active sites were modied to the no blunt end state, as described in Fig.3. (b) The basic scheme of toehold-mediated DNA strand displacement. A target strand invades the duplex to hybridize with the toehold-bearing strand, which allows the two components to separate from each other. (c) FRET readout in dierent congurations showing that under force induction the ring/frameR interface is disrupted only in the presence of soluble target oligonucleotide (*Indicates P0.05).

A self-assembling DNA-based system that is capable of interrogating and, potentially, measuring inter- and intramolecular forces/energies is compelling for several reasons. It is extremely economical in comparison to macroscopic instrumentation that is used for molecular force measurements (e.g., atomic force microscopy (AFM) and optical trapping). Moreover, in contrast to those systems OPTIMuS has the potential to perform thermodynamically reversible force induction, which would overcome the limitations of time-varying external forces obtained by AFM and optical tweezers36. The strength of each force domain may be tuned at the single base pair level to create a highly nuanced spectrum of test energies. Finally, this system may lend itself to statistically robust soluble molecular population-based as well as chip-based single molecule or smaller population analyses.

As a sensor, OPTIMuS is readily recongurable and capable of multiplexing, a potential advantage over molecular beacons37. Unlike one-dimensional DNA-based sensors, OPTIMuS allows the ability to use bulk FRET, corroborated by TEM output to detect molecular states. DNA is amenable to a wide range of chemical modications making it relatively simple to incorporate a variety of molecular species into the system for study. Versions of self-assembling systems like OPTIMuS can be multiplexed to create (AND/OR) logic gates and iterative bio-sensors for high condence molecular detection. Moreover, the ability to precisely arrange gold nanoparticles on OPTIMuS38,39 suggests a pathway to enhanced sensitivity by methods such as surface-enhanced Raman spectroscopy (SERS) for use in eld deployable diagnostics40. Finally, DNA nanosystems are inherently biocompatible and may be further embellished to create novel bionanodevices that have the potential to interact with natural biological systems in vivo.

Methods

Nucleic acids. All oligonucleotide staple strands were purchased from Integrated DNA Technologies (IDT, Coralville, IA), supplied in RNase-free water at 100 M concentration in individual wells. M13mp18 single-stranded scaold DNA was purchased from Bayou Biolabs (Matairie, LA) and was supplied at a concentration of 1g/L in Tris-Acetate EDTA buer. Experiments were carried out without additional purication steps.

Chemical and supplies. All other chemicals (Tris-Acetate EDTA, Magnesium Acetate Tetrahydrate, and water) and supplies were purchased from Fisher Scientic.

Assembly of OPTIMuS. The annealing protocol was adopted from Stein et al.41. The requisite staple strands (including the uorescently-labeled staples), each at a nal concentration of 50nM, were mixed with m13mp18 scaold strand at a nal concentration of 10nM in 1 reaction buer (comprised of 40mM Tris-Acetate, 1mM EDTA (pH 8.3) and 18mM Mg2+) and brought to a nal volume of 500L. The desired structures were assembled using the following thermal annealing program:

80C5min80C to 60C80min60C to 25C1200min25C to 4C10min4Cstorage until further experiments.

SCIENTIFIC REPORTS

6

www.nature.com/scientificreports/

Care was taken to maintain all samples in the dark by covering the PCR plate as well as the laboratory tube rack with aluminum foil.

Removal of excess staples, particularly those with uorescent labels, was critical for optimal results and quantication. Excess staples were removed using Amicon Ultra-0.5 mL Centrifugal Filters (50,000 molecular weight cuto (MWCO)). 500L of the reaction mix was poured into a lter column and centrifuged at 14,000g for 5min. The eluate collected in the collection tube was discarded and the lter column was placed back into the collection tube. Centrifugation step was repeated 4 times on the same lter column by adding 450L of 1 reaction buer to the lter column before each step. Aer completing the centrifugation, the retentate was recovered by inverting the column in a fresh tube and performing centrifugation (at 1000g) for 3min.

Agarose gel electrophoresis. The efficiency of assembly was evaluated by electrophoresis using a 1.5% agarose gel. Electrophoresis was carried out on ice at 72Volts for 4hours. Gels were stained with 1 SYBR Green and illuminated under UV (302nm) using a Benchtop 2UVTM Transilluminator (UV Products).

Post-assembly sample treatment with different combinations of force strands and target strands. In experiments in which a precongured OPTIMuS sample was treated with force domain strands or target strands (Figs3b and 5c), ltered samples were mixed with 100nM of each desired staple, such as extended core, cinchers and loops. The buer conditions of the force domain staples were consistent with the ground state sample and the cation concentration of the resultant samples was maintained at 18mM Mg2+. Samples were incubated using the following thermal annealing protocol:

40C60min40C to 25C600min25C to 4C10min4Cstorage until further experiments.

In order to minimize background uorescence, only ltered samples were used for FRET experiments. A custom-designed well was created in lab using a microscope slide and a coverslip to carry out uorescence microscopy through a hyperspectral microscope (Nikon Eclipse TE2000-E) and EXFO X-Cite 120 PC Fluorescence illumination system. Exposure time was set at 50 ms and a 20X objective was used to image using an HQ Wide Green Filter (Excitation Filter: 545/30nm, Dichromatic Mirror: 570 nm, Barrier Filter: 610/75 nm). Data acquired with the hyperspectral microscope for each sample was a two-dimensional array of Wavelength (nm) and Intensity (arbitrary unit). Intensities were normalized by the total intensity (i) received per sample before calculating FRET. FRET was calculated based on equation(1):

= +

FRET (%) (icy5 100)/(icy3 icy5) (1)

where, icy3 uorescence intensity at cy3 emission (574 nm); icy5 uorescence intensity at cy5 emission (669nm).

Sample preparation for TEM imaging was based on the protocol described by Castro et al.42. Briey, 12 L of the sample solutions (~10 nM concentration) were placed on glow-discharged carbon-coated 400 mesh copper TEM grids. Aer 2min, the samples were wicked o of the grid with lter paper and immediately replaced with 12L of freshly prepared uranyl formate negative staining solution. Aer 30sec, the stain was removed and the grids were allowed to air dry. Images were acquired at 25,000x using a JEOL 1230 TEM (Peabody, MA) equipped with a Gatan Inc. 2k 2k Ultrascan camera (Pleasanton, CA).

Image processing. TEM imaging generated .dm3 les, which were fed into the boxer.py program of EMAN2 to pick individual particles (or nanostructures) and create a stack. Then we performed dimensional analysis (length measurement) using the line tool in ImageJ or Fiji. Scale of the images was set according to the information stored in the .dm3 les. Histograms were generated with the help of a simple R code.

References

1. Zhang, F., Nangreave, J., Liu, Y. & Yan, H. Structural DNA nanotechnology: state of the art and future perspective. J Am Chem Soc 136, 1119811211, doi: 10.1021/ja505101a (2014).

2. Rothemund, P. W. Folding DNA to create nanoscale shapes and patterns. Nature 440, 297302, doi: 10.1038/nature04586 (2006).3. Mathur, D. & Henderson, E. R. Complex DNA nanostructures from oligonucleotide ensembles. ACS Synth Biol 2, 180185, doi: 10.1021/sb3000518 (2013).4. Wei, B., Dai, M. J. & Yin, P. Complex shapes self-assembled from single-stranded DNA tiles. Nature 485, 623+, doi: 10.1038/ Nature11075 (2012).

5. Han, D. et al. DNA origami with complex curvatures in three-dimensional space. Science 332, 342346, doi: 10.1126/ science.1202998 (2011).

6. Ellis, S. J. et al. In Proceedings of the 29th ACM/IEEE international conference on Automated soware engineering 767778 (ACM, 2014).

7. Watada, J. & Abu Bakar, R. B. DNA Computing and Its Applications. Isda 2008: Eighth International Conference on Intelligent Systems Design and Applications, Vol 2, Proceedings, 288-294, doi: 10.1109/Isda.2008.362 (2008).

8. Ezziane, Z. DNA computing: applications and challenges. Nanotechnology 17, R27R39, doi: 10.1088/0957-4484/17/2/R01 (2006).9. Endo, M., Katsuda, Y., Hidaka, K. & Sugiyama, H. Regulation of DNA methylation using dierent tensions of double strands constructed in a dened DNA nanostructure. J Am Chem Soc 132, 15921597, doi: 10.1021/ja907649w (2010).

10. Koirala, D. et al. Single-molecule mechanochemical sensing using DNA origami nanostructures. Angew Chem Int Ed Engl 53, 81378141, doi: 10.1002/anie.201404043 (2014).

SCIENTIFIC REPORTS

7

www.nature.com/scientificreports/

11. Douglas, S. M., Bachelet, I. & Church, G. M. A logic-gated nanorobot for targeted transport of molecular payloads. Science 335, 831834, doi: 10.1126/science.1214081 (2012).

12. Hahn, J., Wickham, S. F., Shih, W. M. & Perrault, S. D. Addressing the instability of DNA nanostructures in tissue culture. ACS Nano 8, 87658775, doi: 10.1021/nn503513p (2014).

13. Mei, Q. et al. Stability of DNA origami nanoarrays in cell lysate. Nano Lett 11, 14771482, doi: 10.1021/nl1040836 (2011).14. Marras, A. E., Zhou, L. F., Su, H. J. & Castro, C. E. Programmable motion of DNA origami mechanisms. Proceedings of the National Academy of Sciences of the United States of America 112, 713718, doi: 10.1073/Pnas.1408869112 (2015).

15. Sacca, B. et al. Reversible Reconguration of DNA Origami Nanochambers Monitored by Single-Molecule FRET. Angew Chem Int Ed Engl 54, 35923597, doi: 10.1002/anie.201408941 (2015).

16. Amir, Y. et al. Universal computing by DNA origami robots in a living animal. Nat Nanotechnol 9, 353357, doi: 10.1038/ nnano.2014.58 (2014).

17. Kershner, R. J. et al. Placement and orientation of individual DNA shapes on lithographically patterned surfaces. Nat Nanotechnol 4, 557561, doi: 10.1038/nnano.2009.220 (2009).

18. Lanier, L. A. & Bermudez, H. DNA nanostructures: a shi from assembly to applications. Curr Opin Chem Eng 7, 93100, doi: 10.1016/j.coche.2015.01.001 (2015).19. Chen, Y. J., Groves, B., Muscat, R. A. & Seelig, G. DNA nanotechnology from the test tube to the cell. Nat Nanotechnol 10, 748760, doi: 10.1038/nnano.2015.195 (2015).

20. Andersen, E. S. et al. Self-assembly of a nanoscale DNA box with a controllable lid. Nature 459, 73U75, doi: 10.1038/Nature07971 (2009).

21. Banerjee, A. et al. Controlled release of encapsulated cargo from a DNA icosahedron using a chemical trigger. Angew Chem Int Ed Engl 52, 68546857, doi: 10.1002/anie.201302759 (2013).

22. Castro, C. E., Su, H. J., Marras, A. E., Zhou, L. & Johnson, J. Mechanical design of DNA nanostructures. Nanoscale 7, 59135921, doi: 10.1039/c4nr07153k (2015).23. Gu, H., Yang, W. & Seeman, N. C. DNA scissors device used to measure MutS binding to DNA mis-pairs. J Am Chem Soc 132, 43524357, doi: 10.1021/ja910188p (2010).

24. Kuzuya, A., Sakai, Y., Yamazaki, T., Xu, Y. & Komiyama, M. Nanomechanical DNA origami single-molecule beacons directly imaged by atomic force microscopy. Nature Communications 2, doi: 10.1038/Ncomms1452 (2011).

25. Yurke, B., Turbereld, A. J., Mills, A. P., Simmel, F. C. & Neumann, J. L. A DNA-fuelled molecular machine made of DNA. Nature 406, 605608 (2000).

26. Zadegan, R. M. et al. Construction of a 4 zeptoliters switchable 3D DNA box origami. ACS Nano 6, 1005010053, doi: 10.1021/ nn303767b (2012).

27. Wang, K. et al. Molecular engineering of DNA: molecular beacons. Angew Chem Int Ed Engl 48, 856870, doi: 10.1002/ anie.200800370 (2009).

28. McIntosh, D. B., Duggan, G., Gouil, Q. & Saleh, O. A. Sequence-dependent elasticity and electrostatics of single-stranded DNA: signatures of base-stacking. Biophys J 106, 659666, doi: 10.1016/j.bpj.2013.12.018 (2014).

29. Brinkers, S., Dietrich, H. R. C., de Groote, F. H., Young, I. T. & Rieger, B. The persistence length of double stranded DNA determined using dark eld tethered particle motion. Journal of Chemical Physics 130, doi: 10.1063/1.3142699 (2009).

30. Singh-Zocchi, M., Dixit, S., Ivanov, V. & Zocchi, G. Single-molecule detection of DNA hybridization. Proc Natl Acad Sci USA 100, 76057610, doi: 10.1073/pnas.1337215100 (2003).

31. Sonnichsen, C., Reinhard, B. M., Liphardt, J. & Alivisatos, A. P. A molecular ruler based on plasmon coupling of single gold and silver nanoparticles. Nat Biotechnol 23, 741745, doi: 10.1038/nbt1100 (2005).

32. Gerling, T., Wagenbauer, K. F., Neuner, A. M. & Dietz, H. Dynamic DNA devices and assemblies formed by shape-complementary, non-base pairing 3D components. Science 347, 14461452, doi: 10.1126/science.aaa5372 (2015).

33. Woo, S. & Rothemund, P. W. K. Programmable molecular recognition based on the geometry of DNA nanostructures. Nature Chemistry 3, 620627, doi: 10.1038/Nchem.1070 (2011).

34. Protozanova, E., Yakovchuk, P. & Frank-Kamenetskii, M. D. Stacked-unstacked equilibrium at the nick site of DNA. Journal of Molecular Biology 342, 775785, doi: 10.1016/J.Jmb.2004.07.075 (2004).

35. Hunter, C. A. Sequence-dependent DNA structure. The role of base stacking interactions. J Mol Biol 230, 10251054, doi: 10.1006/ jmbi.1993.1217 (1993).

36. Neuman, K. C. & Nagy, A. Single-molecule force spectroscopy: optical tweezers, magnetic tweezers and atomic force microscopy. Nat Methods 5, 491505, doi: 10.1038/nmeth.1218 (2008).

37. Pei, H. et al. Scaolded biosensors with designed DNA nanostructures. NPG Asia Materials 5, 9 (2013).38. Tan, S. J., Campolongo, M. J., Luo, D. & Cheng, W. Building plasmonic nanostructures with DNA. Nat Nanotechnol 6, 268276, doi: 10.1038/nnano.2011.49 (2011).39. Ding, B. et al. Gold nanoparticle self-similar chain structure organized by DNA origami. J Am Chem Soc 132, 32483249, doi: 10.1021/ja9101198 (2010).40. Acuna, G. P. et al. Fluorescence Enhancement at Docking Sites of DNA-Directed Self-Assembled Nanoantennas. Science 338,

506510, doi: 10.1126/Science.1228638 (2012).

41. Stein, I. H., Schuller, V., Bohm, P., Tinnefeld, P. & Liedl, T. Single-Molecule FRET Ruler Based on Rigid DNA Origami Blocks. Chemphyschem 12, 689695, doi: 10.1002/Cphc.201000781 (2011).

42. Castro, C. E. et al. A primer to scaolded DNA origami. Nat Methods 8, 221229, doi: 10.1038/nmeth.1570 (2011).

Acknowledgements

This project was funded by the Bill and Melinda Gates Foundation Grand Challenges Exploration Grant No. OPP1059568, NSF Inspire Grant Nos 1247051 and 1545028, and personal funds. The authors would like to acknowledge the Roy J. Carver Laboratory for Ultrahigh Resolution Biological Microscopy at Iowa State University for their facilities and sincerely thank Thomas Moninger and the Central Microscopy Research Facility at University of Iowa to provide TEM data.

Author Contributions

D.M. and E.R.H. designed the experiments, analyzed the data and co-wrote the paper.

Additional Information

Supplementary information accompanies this paper at http://www.nature.com/srep

Competing nancial interests: E.R.H. has led a provisional patent application on elements of the nanosystem described in this report.

SCIENTIFIC REPORTS

8

www.nature.com/scientificreports/

How to cite this article: Mathur, D. and Henderson, E. R. Programmable DNA Nanosystem for Molecular

Interrogation. Sci. Rep. 6, 27413; doi: 10.1038/srep27413 (2016).

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

SCIENTIFIC REPORTS

9