The design of a mechanical wave-like DNA nanomachine for the fabrication of a programmable and multifunctional molecular device

Xiaoli Zhu a, Xiaoxia Chen a, Fangfang Ban a, Ya Cao a, Jing Zhao a, Guifang Chen *a and Genxi Li *ab
aCenter for Molecular Recognition and Biosensing, School of Life Sciences, Shanghai University, Shanghai 200444, China. E-mail:
bState Key Laboratory of Pharmaceutical Biotechnology and Collaborative Innovation Center of Chemistry for Life Sciences, Department of Biochemistry, Nanjing University, Nanjing 210093, China. E-mail:

Received 5th July 2017 , Accepted 10th August 2017

First published on 10th August 2017

We report a novel mechanical wave-like DNA nanomachine. A successive stem-loop structure is involved, which will rearrange successively from one side to the opposite side upon binding with an input activator just like the motion of a mechanical wave, and thereby achieving diverse functions.

DNA plays a central role in carrying genetic information and translating the genetic code into proteins. Besides, owing to its easily predictable secondary structure and programmability, DNA has become the material of choice for the construction of complex nanometer-scale molecular structures in nanotechnology.1 Recently, a growing and exciting field of DNA nanotechnology emerged, where synthetic DNA strands are used to build responsive nanomachines or functional nanodevices through rational design and engineering.2 Several elaborately designed DNA structures, like “tweezers”,3 “walkers”,4 “steppers”,5 “switches”,6 have been reported to perform different functions.

Currently, great attention is being paid to the design and exploration of DNA nanomachines for amplifying bioanalysis.7 In the attempts made, the binding events of the target were converted to readable output signals, e.g. fluorescence signals of molecular beacons. A few of other potential functions, e.g. regulation of cell activities and delivery of therapeutic compounds, have also been revealed.8 Despite the success in the development of DNA nanomachines for different purposes, several limitations apply, however, making them still far from application in the real-world: (a) most of the DNA nanomachines lack diverse input and output signals, restricting their versatile applications; (b) irreversible switching of the DNA nanomachines usually occurs, thus eliminating the cyclic operation and (c) most of the DNA nanomachines consist of several or even dozens of DNA strands, making it difficult to ensure the intact structure of the nanomachines within their dynamic operation.

Here, inspired by the mechanical wave through a rope held by a rhythmic gymnast, we assume that the aesthetic linear structure of a DNA strand can work like the rope to transfer and transform something at the molecular level. Thus, we report a novel kind of DNA nanomachine named as a mechanical wave-like DNA nanomachine that can couple diverse inputs into diverse outputs through a single reversible DNA strand. Diverse functions including bioanalysis, regulation of enzymatic activity, controllable synthesis of nanomaterials are thereby revealed.

As shown in Scheme 1, a single-stranded sequence containing a continuous stem-loop structure is involved in our mechanical wave-like DNA nanomachine. For the simplest case that contains only one stem-loop, this nanomachine is activated through allosteric tuning. Upon binding with an input, e.g. a complementary single-stranded oligonucleotide, through a toehold, the conformation of the sequence switches: half of the stem switches to a new loop, whereas the loop switches to a new stem, that is, a new stem-loop structure is formed. This behavior is just like the switch of a single wave from sine to cosine, in which the loop is analogous to the peak, and the stem is analogous to the slope. The allosteric tuning relies strongly on the switching equilibrium, which can be elucidated using an equation (KD: intrinsic affinity of the binding-competent state; Ks: switching equilibrium constant; KD-obs: overall observed affinity):9

image file: c7cc05174c-t1.tif(1)

image file: c7cc05174c-s1.tif
Scheme 1 Scheme of the wave-like DNA nanomachines. Input: an oligonucleotide, output: fluorescence signals produced from molecular beacons.

Although the above single-wave like DNA nanomachine may work and has been reported elsewhere,10 a big limitation exists: the stem and loop as well as the toehold correlate with each other, meaning that there is not enough space and functional DNA to transform an input signal to diverse outputs or diverse inputs to a specific output. This problem can be resolved by linking a related stem-loop into a continuous head-to-tail stem-loop structure. For the double or more wave-like DNA nanomachine that contains two or more stem-loops, the input oligonucleotide triggers the allosteric switch of the first stem-loop into a new one, which subsequently switches the second stem-loop, and then the third, and so on. This behavior is similar to the propagation of a mechanical wave with a shift of 1/4 wavelength. It should be noted here that the output region (3′-terminal, right side) is totally separate from the input region (5′-terminal, left side). Thus, it is expected that the complex function of the DNA nanomachine can be realized through rational designs.

We first adopt a molecular beacon to study the transformation of the binding input of a single-stranded oligonucleotide (“input DNA” for short) through the mechanical wave-like DNA nanomachine to readable fluorescence output signals. Qualitative analysis using electrophoresis is conducted. The results show that conformational changes and fluorescence output signals of the wave-like DNA can be launched upon binding with input DNAs, just as expected (Fig. S1, ESI). Dynamic studies as well as the impact of external conditions on the signal transformation using fluorescence spectrometry were then conducted. With increasing concentration of the input DNA, the fluorescence signal increases and then reaches a plateau (Fig. 1a, c and e). The half maximal effective concentrations (EC50) of the input DNA are all around 700 nM in the case of 1 μM wave DNA. The results suggest a highly efficient and stable transforming effect especially for double- and triple-waves, which need multi-strand displacement reactions. With an excess concentration of input DNA, it is observed from Fig. 1b, d and f that the signal transformation (including the hybridization of input DNA to wave DNA and the strand displacement inside the wave DNA) is fast. Upon the addition of the input with mixing (red dots), a remarkable signal can be obtained immediately (ca. 20 s is required to obtain the data points from the addition). The fluorescence signal becomes stable in about 5 min. Otherwise, if the input DNA is added without mixing, plots with two rising stages, corresponding to the diffusion and signal transformation, can be observed. In both cases, there is no apparent difference among the three wave DNAs, suggesting the hybridization but not the strand displacement is the rate-limiting step. The impact of the salt concentration and temperature on the fluorescence signal transformation were also studied (Fig. S2 and S3, ESI).

image file: c7cc05174c-f1.tif
Fig. 1 Fluorescence outputs upon the addition of input oligonucleotides. (a, c and e) Fluorescence outputs vs. the concentration of the inputs: (a) single wave, (c) double wave, and (e) triple wave. Insets: The corresponding black and white photographs under UV-light. (b, d and f) Fluorescence outputs vs. the reaction time: (b) single wave, (d) double wave, and (f) triple wave. Red round dots: with sufficient mixing; black square dots: without mixing.

The specificity of the recognition between the input and the wave DNA is important for analytical applications. As shown in Fig. 2a, the base-pairing between these two DNAs can be divided into two regions: an anchor region and a displacement region. Point mutations of the input DNA in both regions are adopted to study the specificity of the wave DNA nanomachines with different lengths. From Fig. 2b, c and Fig. S4 (ESI), some phenomena can be observed and a few conclusions can be drawn: (1) point mutations on either the anchor region or the displacement region result in the decrease of the output, suggesting the equal importance of these two regions in the binding efficiency and consequently the fluorescence signal for the specificity; (2) because of the stronger interaction between C and G through three hydrogen bonds, point mutations of C/G to A/T donate more to the specificity than A/T to C/G; and (3) it is interesting to note that the specificity of the double-wave and triple-wave is much better than the single-wave. Because the only difference between the single-wave and the multi-wave is the excess wave-like stem-loop structure other than the input binding region, it can be concluded that the switch of the wave-like stem-loop inside the wave DNA upon binding with the input helps a lot with the improved specificity.

image file: c7cc05174c-f2.tif
Fig. 2 Specificity for the input oligonucleotides. (a) Schematic representation of the mutants of the input oligonucleotides for single wave (W1) and double wave (W2). (b and c) Fluorescence outputs of different mutant inputs for (b) single wave and (c) double wave.

Other than the transformation of a single-stranded oligonucleotide to fluorescence signals, four other kinds of signal transformations are explored, in which both the input and the output can vary. In the case of the input, miRNA and Hg2+ ions are adopted to induce fluorescence signals through the wave-like DNA nanomachine. While for the output, the regulation of enzymatic activity and the synthesis of DNA-templated nanoclusters are achieved, both of which are triggered by a single-stranded oligonucleotide. In order to unify the system, only double-wave DNA is adopted to achieve the above transformation.

For the transformation of miRNA to fluorescence signals, let-7a, a miRNA that is lowly expressed in cancer cells,11 is adopted. As shown in Fig. 3, the reaction time-, concentration- and sequence-depended performances of the miRNA-triggered DNA nanomachine are studied. The results are similar to those of the single-stranded oligonucleotide input shown in Fig. 1 and 2, suggesting that the wave-like DNA nanomachine can be also applied for the miRNA input. From Fig. 3c, a linear regression can be also obtained in a miRNA concentration ranging from 1 nM to 800 nM, with a detection limit of 0.33 nM (LOD = 3SD/k, LOD: detection limit, SD: the standard deviation of the blank sample, and k: the slope of the fitting curve). Thus, the successful transformation also provides a way for the fluorescence detection of this potential tumor biomarker. In this attempt, the length of the anchor region of double-wave DNA is retained at 6 nt, similar to that of the single-stranded oligonucleotide input (Scheme 1), whereas the displacement region is extended to 16 nt, because of the longer sequence of miRNA. If the displacement region is retained at 10 nt whereas the anchor region is extended to 12 nt, it is interesting that the time required for transformation becomes shorter and the specificity is no longer good enough (Fig. S5, ESI). Here, the longer anchor region facilitates the anchor of the miRNA onto the wave DNA, thus shortening the transformation time. The results also confirm that the anchor i.e. hybridization but not the strand displacement is the rate-limiting step. The extended anchor region otherwise weakens the specificity because of the more base-pairing sites for the mutations of the target let-7a. Nevertheless, the above results show that the wave-like DNA nanomachine offers tenability for varied needs. In addition to the miRNA input, successful transformation from the ionic input (Hg2+ ion) to fluorescence signals is also achieved by using another well-designed wave-like DNA nanomachine (Fig. S6, ESI).

image file: c7cc05174c-f3.tif
Fig. 3 From miRNA inputs to fluorescence outputs: case 1. (a) Schematic presentation. Let-7a was adopted as the input, while other members of the let-7 family were adopted as mutant controls. (b) Fluorescence outputs vs. the reaction time. Red round dots: with sufficient mixing; black square dots: without mixing. (c) Fluorescence outputs vs. the concentration of the input. Insets: The corresponding black and white photographs under UV-light and a linear range from 1 nm to 800 nm. (d) Fluorescence outputs upon the addition of the input or other mutants.

The transformation of biomolecules through the wave DNA to fluorescence signals provides an opportunity for bioanalysis. It has some advantages: (1) quick response (5 min) through binding-induced conformational changes; (2) solid outputs within a single strand; and (3) availability for diverse targets because of the spatial and functional separations between the input side and the output side of the nanomachine. But currently because of the lack of signal amplification, the sensitivity is not satisfactory compared with some other methods.12 By integrating with signal amplification units, the performance is expected to be further improved.

The possibility for diverse outputs is then studied. First, we integrate an anti-thrombin aptamer into the output region of a double-wave DNA to study the possibility for the regulation of enzymatic activity. Upon binding with a single-stranded oligonucleotide input, the conformational change of the wave DNA allows the exposure of the aptamer, which can further bind thrombin and inhibit its activity (Fig. 4a). A conventional thrombin time (TT) assay that measures the time for a clot to form in the plasma of a blood sample is adopted to investigate the activity of thrombin. As shown in Fig. 4b, with increasing the amount of thrombin in the plasma from 10 nM to 1000 nM, the TT is shortened from 292 s to 53 s. A critical concentration of 200 nM is adopted for the following experiments. The concentration of wave DNA is also set at 200 nM to avoid unwanted inhibition of the enzyme in the presence of excess inactive wave DNA (Fig. 4C). After the condition optimization, different amounts of the oligonucleotide input are introduced to trigger the wave-like DNA nanomachine and regulate the activity of thrombin subsequently. As shown in Fig. 4d, with increasing the amount of the input, the TT increases from 67 s to 131 s. Together with the results shown in Fig. 4b, it can be concluded that a maximum of ca. 79% of the thrombin activity is inhibited under the regulation of the input-triggered DNA nanomachine. Further results by using the wave DNA or oligonucleotide input alone to regulate thrombin validate the combined action of the input and the wave DNA (Fig. 4e). So, here we show that, aided by the wave-like DNA nanomachine, an unrelated input molecule can be applied for the regulation of the activity of a specific protein. It also has the prospect for the fabrication of some unconventional smart systems, for example, self-regulated drug release by using a self-overexpressed biomarker as the input to inhibit a critical protein through the DNA nanomachine. In addition to the regulation of enzymatic activity, the successful synthesis of AgNCs under regulation is also achieved (Fig. S7 and S8, ESI).

image file: c7cc05174c-f4.tif
Fig. 4 From oligonucleotide to the regulation of enzymatic activity. (a) Schematic presentation. A random input oligonucleotide was adopted to regulate the activity of thrombin. (b) Thrombin time vs. the concentration of thrombin. (c) Thrombin time vs. the concentration of wave DNA in the presence of 200 nM thrombin. (d) Thrombin time vs. the concentration of the input oligonucleotide in the presence of 200 nM wave DNA together with 200 nM thrombin. (e) Thrombin time in the case of thrombin alone (black column), thrombin with wave DNA (red column), thrombin with input oligonucleotide (green column), and thrombin with wave DNA and input oligonucleotide (blue column). The percentages above the columns show the inhibitory rates on the enzymatic activity.

In summary, a novel mechanical wave-like DNA nanomachine that consists of a single-stranded DNA with a successive stem-loop structure is developed. This DNA nanomachine upon binding with a specific input molecule undergoes binding-induced conformational changes: a mechanical motion inside the DNA sequence and a rearrangement of the stem-loop structure, which can be consequently conferred to useful functions. In this work, the transformation of diverse inputs (including single-stranded oligonucleotides, miRNA and ions) into diverse outputs (including florescence signals, regulation of enzymatic activity and controllable synthesis of nanoclusters) is excitingly achieved. The integrated single-stranded nanostructure as well as the unique mechanical motion of this wave-like DNA nanomachine has some distinct advantages, making it promising for the expansion of the available toolbox in the field of DNA nanotechnology.

This work was supported by the National Natural Science Foundation of China (Grant No. 21575088, 21235003, and 31200742) and the Natural Science Foundation of Shanghai (14ZR1416500).

Conflicts of interest

There are no conflicts to declare.

Notes and references

  1. Y. Krishnan and F. C. Simmel, Angew. Chem., Int. Ed., 2011, 50, 3124–3156 CrossRef CAS PubMed; A. V. Pinheiro, D. Han, W. M. Shih and H. Yan, Nat. Nanotechnol., 2011, 6, 763–772 CrossRef PubMed; F. C. Simmel, Angew. Chem., Int. Ed., 2008, 47, 5884–5887 CrossRef PubMed.
  2. R. M. Zadegan and M. L. Norton, Int. J. Mol. Sci., 2012, 13, 7149–7162 CrossRef CAS PubMed; F. Zhang, J. Nangreave, Y. Liu and H. Yan, J. Am. Chem. Soc., 2014, 136, 11198–11211 CrossRef PubMed.
  3. X.-Y. Li, J. Huang, H.-X. Jiang, Y.-C. Du, G.-M. Han and D.-M. Kong, RSC Adv., 2016, 6, 38315–38320 RSC; M. Liu, J. Fu, C. Hejesen, Y. Yang, N. W. Woodbury, K. Gothelf, Y. Liu and H. Yan, Nat. Commun., 2013, 4, 2127 Search PubMed.
  4. T. E. Ouldridge, R. L. Hoare, A. A. Louis, J. P. K. Doye, J. Bath and A. J. Turberfield, ACS Nano, 2013, 7, 2479–2490 CrossRef CAS PubMed; C. Wang, J. Ren and X. Qu, Chem. Commun., 2011, 47, 1428–1430 RSC; L. Wang, R. Deng and J. Li, Chem. Sci., 2015, 6, 6777–6782 RSC.
  5. Z. G. Wang, J. Elbaz and I. Willner, Nano Lett., 2011, 11, 304–309 CrossRef CAS PubMed.
  6. A. Porchetta, A. Vallee-Belisle, K. W. Plaxco and F. Ricci, J. Am. Chem. Soc., 2013, 135, 13238–13241 CrossRef CAS PubMed; A. Idili, A. Vallee-Belisle and F. Ricci, J. Am. Chem. Soc., 2014, 136, 5836–5839 CrossRef PubMed; S. Ranallo, A. Amodio, A. Idili, A. Porchetta and F. Ricci, Chem. Sci., 2016, 7, 66–71 RSC.
  7. Q. Xue, Y. Lv, H. Cui, X. Gu, S. Zhang and J. Liu, Anal. Chim. Acta, 2015, 856, 103–109 CrossRef CAS PubMed.
  8. S. Surana, J. M. Bhat, S. P. Koushika and Y. Krishnan, Nat. Commun., 2011, 2, 340 CrossRef CAS PubMed; C.-H. Lu, B. Willner and I. Willner, ACS Nano, 2013, 7, 8320–8332 CrossRef PubMed.
  9. A. Vallee-Belisle, F. Ricci and K. W. Plaxco, Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 13802–13807 CrossRef PubMed.
  10. M. Rossetti, S. Ranallo, A. Idili, A. Porchetta and F. Ricci, Chem. Sci., 2012, 1–3 Search PubMed; A. J. Simon, A. Vallee-Belisle, F. Ricci, H. M. Watkins and K. W. Plaxco, Angew. Chem., Int. Ed., 2014, 53, 9471–9475 CrossRef CAS PubMed.
  11. J. Takamizawa, H. Konishi, K. Yanagisawa, S. Tomida, H. Osada, H. Endoh, T. Harano, Y. Yatabe, M. Nagino, Y. Nimura, T. Mitsudomi and T. Takahashi, Cancer Res., 2004, 64, 3753–3756 CrossRef CAS PubMed.
  12. T. Tian, J. Wang and X. Zhou, Org. Biomol. Chem., 2015, 13, 2226–2238 Search PubMed; G. Chen, Z. Guo, G. Zeng and L. Tang, Analyst, 2015, 140, 5400–5443 RSC.


Electronic supplementary information (ESI) available: Experimental section, SEM characterization, and condition optimization. See DOI: 10.1039/c7cc05174c

This journal is © The Royal Society of Chemistry 2017