Allosteric Control of Oxidative Catalysis by a DNA Rotaxane

Communication Cite This: J. Am. Chem. Soc. 2017, 139, 16044-16047

pubs.acs.org/JACS

Allosteric Control of Oxidative Catalysis by a DNA Rotaxane Nanostructure Mathias Centola,† Julián Valero,*,†,‡ and Michael Famulok*,†,‡,§ †

LIMES Chemical Biology Unit, Universität Bonn, Gerhard-Domagk-Straße 1, 53121 Bonn, Germany Center of Advanced European Studies and Research, Ludwig-Erhard-Allee 2, 53175 Bonn, Germany § Center of Aptamer Research and Development, University of Bonn, Gerhard-Domagk-Straße 1, 53121 Bonn, Germany ‡

Downloaded via RUTGERS UNIV on January 2, 2019 at 17:23:29 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

Previous reports showed that DNA nanodevices can tune functionality through spatial rearrangements by allosteric effectors.5 Here we present a DNA rotaxane that allosterically regulates the catalytic activity of a split HRP-DNAzyme by switching and precisely control the position of its interlocked components. The ds-axle of the DNA rotaxane contains two single-stranded (ss)-stations, namely position 1 and position 2 (Figure S1) that are complementary to two ss-gap regions of the 126bp macrocycle. Selective threading of the macrocycle to positions 1 or 2 on the axle allows two distinct structural arrangements that can be precisely achieved by using short ssODNs called “release oligos” (ROs). The ROs are designed to specifically compete with the macrocycle for hybridization to positions 1 and 2, independently (Supplementary Movie). In our design, we introduced a 3:1 split HRP-mimicking DNAzyme6 in which the sequence that forms the catalytic core is split in two parts, therefore displaying catalytic activity only when the two halves are linked by a third DNA strand here called connector ODN (C-oligo).6 Notwithstanding, in our system the formation of the catalytic core directly depends on the different rotaxane conformations and not on the C-oligo, which is always present in solution (Figure 1, Tables S1−S4, Figure S1). Thus, structural changes occurring far from the catalytic core are responsible for the allosteric regulation of the system’s activity. The DNAzyme consists of a G-quadruplex (G4) that selfassembles in the presence of K+-ions.6 Iron protoporphyrin IX (hemin) intercalates into the G4 structure, thus enhancing its peroxidase activity which can be monitored by the oxidation of 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) using UV−vis spectrometry.7 To accommodate the split DNAzyme into the DNA rotaxane, a sequence with three G triplets resides on one of the 168bp stopper rings and the remaining G triplets on the 126bp macrocycle. Two independent ss-regions flanking the G4 core serve as hybridization sites for the C-oligo, thus allowing the formation of the G-quadruplex structure. An appropriate connector ss-ODN must fulfill three main conditions: (i) allow proper hybridization between the two halves of the split DNAzyme to recover its peroxidase activity; (ii) avoid the intermolecular assembly of split DNAzyme components from different noninterlocked systems and (iii) prevent the intra-

ABSTRACT: DNA is a versatile construction material for the bottom-up assembly of structures and functional devices in the nanoscale. Additionally, there are specific sequences called DNAzymes that can fold into tertiary structures that display catalytic activity. Here we report the design of an interlocked DNA nanostructure that is able to fine-tune the oxidative catalytic activity of a split DNAzyme in a highly controllable manner. As scaffold, we employed a double-stranded DNA rotaxane for its ability to undergo programmable and predictable conformational changes. Precise regulation of the DNAzyme’s oxidative catalysis can be achieved by external stimuli (i.e., addition of release oligos) that modify the spatial arrangement within the system, without interfering with the catalytic core, similar to structural rearrangements that occur in allosterically controlled enzymes. We show that multiple switching steps between the active and inactive conformations can be performed consistent with efficient regulation and robust control of the DNA nanostructure.

T

he enormous potential of DNA has been proven for the assembly of nanostructures via bottom up approaches. DNA sequences can be programmed to arrange in desired structures following the standard Watson and Crick base pairing rules.1 DNA origami structures can precisely define the positioning of molecules in the second and third dimension.2 On the other hand, mechanically interlocked nanostructures, such as catenanes and rotaxanes, show dynamic behavior by means of their interlocked components that can move with relatively large amplitudes.3 Double-stranded (ds)DNA rotaxanes, for example, provide mechanically interlocked configurations in which the macrocycle can move within the mechanical constrains, and hybridized configurations that stall macrocycle movement at a defined position on the axle. The transition from one macrocycle position to another can be triggered by different stimuli like addition of short oligodeoxynucleotides (ODNs), PNAs, or light induced mechanisms.3b−f,4 An important goal in the field of DNA nanotechnology is the construction of assemblies that perform functions mimicking those of natural counterparts. For example, many natural enzymes, finely regulate their enzymatic reaction by conformational changes in the protein scaffold that occur allosterically at a site that is distant from the catalytic core. © 2017 American Chemical Society

Received: August 18, 2017 Published: October 23, 2017 16044

DOI: 10.1021/jacs.7b08839 J. Am. Chem. Soc. 2017, 139, 16044−16047

Communication

Journal of the American Chemical Society

configurations, and served as a positive control (Figure 2b). Atomic force microscopy (AFM) analysis confirmed that the active conformation of the rotaxane with the macrocycle locked in position 1 occurred intramolecularly (Figure 3a,b, Figures

Figure 1. dsDNA rotaxane-directed formation of a catalytic site. (a) Configuration of the DNA rotaxane. The dumbbell (blue) carries one part of the split HRP-mimicking DNAzyme on the ring stopper next to position 2. The macrocycle (red) contains two single-stranded (ss) gaps, and the remaining part of the DNAzyme as an overhang. Green: C-oligo. Top: Catalytically inactive DNA rotaxane with macrocycle hybridized to position 1. Bottom: Catalytically active conformation with the macrocycle hybridizing to position 2 with the C-oligo allowing the formation of the catalytic G4 structure. (b) Catalytic hemin-G4 complex stabilized by potassium ions.

Figure 3. AFM images and corresponding schematic representations of different DNA rotaxane conformations. (a, b) DNA rotaxanes bearing ring stoppers with the macrocycle in position 1 in the presence (a) or absence (b) of C-positive ODN. (c, d) DNA rotaxanes bearing one spherical stopper with the macrocycle hybridized to position 1 in the presence of C-positive (c) or C10 (d).

molecular interaction of the two DNAzyme halves when the macrocycle is located in position 1 (Figure 2a). Hence, C-oligos of different lengths were screened and tested together with the noninterlocked macrocycle subunits as well as with the DNA rotaxanes bearing the macrocycle either in position 1 or 2, respectively (Figure 2). These conformations are assembled separately, purified by anion exchange HPLC and used to evaluate their peroxidase activity in the presence of the different C-oligos (experimental details in Supporting Information, Figures S2−S5).The longest C-oligo, C-positive, forms the fully active DNAzyme in solution when using the noninterlocked ring components or either of the rotaxane

S6, S7) rather than by dimerization of two rotaxane nanostructures. To better discriminate between the two positions by AFM, we introduced a spherical stopper near position 1 (Figure 3c,d, Tables S5, S6, Figures S6−S12). These nanostructures showed that C-positive bends the axle, thus allowing the intramolecular formation of the HRP-DNAzyme with the macrocycle remaining hybridized in position 1 (Figure 3c, Figure S13). Among all tested C-oligos (Figures S13−S26), C11, C10 and C9, emerged as the most promising. Although these ODNs differ from one another only by one nucleotide in length, they

Figure 2. Design and activity of C-oligos and mechanism of the nanodevice. (a) C-oligo sequences and the split HRP-mimicking DNAzyme attached to the macrocycle and the ring stopper. The target hybridization sites for the different C-oligos are depicted with the dashed lines. (b) Relative peroxidase reaction rates (n ≥ 3) of the noninterlocked subunits, the dsDNA rotaxane with the macrocycle in position 1 and in position 2 in the presence of the respective C-oligo. Details of the reaction rate calculation and error bars are described in the Supporting Information. 16045

DOI: 10.1021/jacs.7b08839 J. Am. Chem. Soc. 2017, 139, 16044−16047

Communication

Journal of the American Chemical Society

Figure 4. Switching cycle of the catalytic nanostructure. (a) Switching cycle of the catalytic nanostructure, from inactive (1) to active conformation (5), illustrating some of the intermediate states (2−4, 6). Details of the switching cycle, triggered by sequential addition of the indicated ROs, are given in the Supporting Information. (b, d) Reaction rates of the DNA rotaxane in the presence of the C10 oligo during the switching. Activity characterization of the active (A) and inactive (I) conformations as well as disassembled 1 (D1) and disassembled 2 (D2) structures used as controls for the macrocycle release, are shown (further details in Figures S27−28). Reaction rates were calculated considering the active state as reference for normalization (n = 3). (c, e) Electrophoretic gel analysis of samples recovered after activity measurements. The rotaxane structure remains stable during the switching, whereas dumbbell formation is observed for the controlled disassembled samples (D1 and D2). Minimal dethreading is also observed in the last switching step.

translocation of the macrocycle from one position to the other is sufficient to disassemble (inactivate) and assemble (activate) the catalytic core. Blockade of the macrocycle’s target position eventually causes its dethreading and disassembly of the interlocked structure. We use this controlled disassembly to monitor the efficient release of the macrocycle from the axle by agarose gel electrophoresis (Figure 4c,e). Because the hybridization of the macrocycle on positions 1 or 2 does not change the electrophoretic mobility of the structure, the controlled disassembly of the system is used to assess dumbbell formation due to dethreading of the macrocycle. The dumbbell structure displays faster electrophoretic mobility than the rotaxane, which can be easily detected. Our experiments confirm that the system can robustly perform a full switching cycle starting from both, position 1 (Figure 4b,c, Table S8) and position 2 (Figure 4d,e, Table S9), and returning to the original conformational state. We further evaluated the catalytic activity of each structural intermediate involved in the switching cycle. A decrease in peroxidase activity is evident when switching from the active to the inactive or disassembled states. Moreover, the system fully recovers its activity after one full cycle (Figure 4b,d). Small discrepancies regarding the differences in activity between the inactive (position 1) and active (position 2) forms, depending on the initial state can be attributed to the purity of the samples, incomplete switching, marginal disassembly and most importantly, accumulation of chemical waste, i.e., ROs and cROs. These results indicate that controlling the spatial arrangement of the interlocked subunits forming the rotaxane nanostructure is sufficient to modulate the catalytic activity of these nanomachines. Fine-tuning of the C-oligos and their strength of hybridization is required to avoid undesired interand intramolecular interactions, which may interfere with its structural and functional control. Overall, a delicate balance

distinctively influenced the nanostructure activity. C11 behaved similarly as C-positive despite its significantly shorter hybridization sequence. Interestingly, C10, with only one nucleotide less than C11, showed drastically reduced activity in the noninterlocked system and in the DNA rotaxane with the macrocycle hybridized to position 1. However, C10 displayed a catalytic activity comparable to the positive control with the macrocycle hybridized to position 2. When reducing the hybridization length by another nucleotide in C9, we observed a further decrease of activity in the noninterlocked system and in the DNA rotaxane with the ring hybridized to position 1. The reaction rate in the presence of C9 increased significantly when the ring hybridized to positon 2, but was only approximately half as active as the positive control (Figure 2). Overall, C10 satisfies the system’s requirements described previously and displays the highest differences in the catalytic activity of the different configurations of the DNA rotaxane. Therefore, it was selected for optimization and quantification of switching efficiencies. We next investigated whether the nanodevice can switch reversibly between the conformations defined by the macrocycle position. We used the nonsymmetric rotaxane with the spherical stopper near position 1 and the ring stopper near position 2. The spherical stopper significantly reduces the dethreading of the macrocycle thus enhancing the stability of the mechanically interlocked components.3a,d First, the release oligo (RO) blocking the target position was removed by a complementary ODN via a toehold-mechanism (cRO, Table S7) making the target position available for hybridization to the macrocycle. Addition of a specific RO releases the macrocycle from its current hybridization position. It can now freely move along the axle until it reaches the new accessible station (Figure 4a). Importantly, the C-oligo is present during the entire switching cycle and therefore 16046

DOI: 10.1021/jacs.7b08839 J. Am. Chem. Soc. 2017, 139, 16044−16047

Journal of the American Chemical Society

■

between the distance separating both macrocycle stations, the stiffness of the rotaxane axle, and the hybridization energy of the connector ODN, contribute to the control of the nanodevice’s reactivity. This modulation of the catalytic activity can be compared to allosteric regulation in enzymes where the conformational changes in regions distant from the catalytic core greatly influence the enzyme’s activity. The ROs can be considered as equivalents to allosteric interactors because they change the conformation of the system by switching the position of the macrocycle on the axle, therefore modulating the reactivity of the nanostructure. In conclusion, here we present the design and assembly of a dsDNA rotaxane containing a split horseradish peroxidase DNAzyme, the catalytic activity of which depends on the position of the interlocked macrocycle on the nanostructure. Toehold mediated strand displacement reactions were used to control the position of the interlocked macrocycle and thereby the catalytic activity of the system. Our DNA nanostructure is advantageous to other reported synthetic organic interlocked systems or DNA nanoswitches displaying limited reversibility or that cannot efficiently be switched sequentially.5a,b,7 The catalytic function of previous DNA-based topologically complex systems was regulated by external stimuli that directly influenced the catalytic core.4c,8 Our system harnesses release oligos (ROs) as positive and negative allosteric effectors that cause conformational changes in the DNA rotaxane structure by virtue of macrocycle mobility, without directly affecting the catalytic subunit. This strategy offers an additional degree of control in functional DNA architectures, thus mimicking allosterically regulated enzymes. It is anticipated that the interlocked DNA nanostructure described here can be combined with other catalytically active split macromolecules in the future. Its straightforward implementation into other more complex functional nanodevices will pave the way for the construction of sophisticated artificial multicatalytic systems with controlled and enhanced reactivity such as nanofactories, switches and motors.

■

REFERENCES

(1) (a) Seeman, N. C. Nature 2003, 421 (6921), 427−431. (b) Rothemund, P. W. K. Nature 2006, 440 (7082), 297−302. (c) Zhang, F.; Nangreave, J.; Liu, Y.; Yan, H. J. Am. Chem. Soc. 2014, 136 (32), 11198−211. (d) Torring, T.; Voigt, N. V.; Nangreave, J.; Yan, H.; Gothelf, K. V. Chem. Soc. Rev. 2011, 40 (12), 5636−46. (e) Nangreave, J.; Han, D.; Liu, Y.; Yan, H. Curr. Opin. Chem. Biol. 2010, 14 (5), 608−15. (2) Funke, J. J.; Dietz, H. Nat. Nanotechnol. 2016, 11 (1), 47−52. (3) (a) Ackermann, D.; Schmidt, T. L.; Hannam, J. S.; Purohit, C. S.; Heckel, A.; Famulok, M. Nat. Nanotechnol. 2010, 5, 436−442. (b) Lohmann, F.; Ackermann, D.; Famulok, M. J. Am. Chem. Soc. 2012, 134 (29), 11884−7. (c) Lohmann, F.; Weigandt, J.; Valero, J.; Famulok, M. Angew. Chem., Int. Ed. 2014, 53 (39), 10372−10376. (d) Weigandt, J.; Chung, C.-L.; Jester, S.-S.; Famulok, M. Angew. Chem., Int. Ed. 2016, 55, 5512−5516. (e) Powell, J. T.; Akhuetie-Oni, B. O.; Zhang, Z.; Lin, C. Angew. Chem., Int. Ed. 2016, 55 (38), 11412− 6. (f) List, J.; Falgenhauer, E.; Kopperger, E.; Pardatscher, G.; Simmel, F. C. Nat. Commun. 2016, 7, 12414. (g) Elbaz, J.; Cecconello, A.; Fan, Z.; Govorov, A. O.; Willner, I. Nat. Commun. 2013, 4, 2000. (h) Elbaz, J.; Wang, Z. G.; Wang, F.; Willner, I. Angew. Chem., Int. Ed. 2012, 51 (10), 2349−53. (i) Lu, C. H.; Cecconello, A.; Qi, X. J.; Wu, N.; Jester, S. S.; Famulok, M.; Matthies, M.; Schmidt, T. L.; Willner, I. Nano Lett. 2015, 15 (10), 7133−7. (j) Lu, C. H.; Qi, X. J.; Cecconello, A.; Jester, S. S.; Famulok, M.; Willner, I. Angew. Chem., Int. Ed. 2014, 53 (29), 7499−503. (k) Cecconello, A.; Lu, C. H.; Elbaz, J.; Willner, I. Nano Lett. 2013, 13 (12), 6275−80. (4) (a) Ackermann, D.; Famulok, M. Nucleic Acids Res. 2013, 41 (8), 4729−39. (b) Ackermann, D.; Jester, S. S.; Famulok, M. Angew. Chem., Int. Ed. 2012, 51 (27), 6771−5. (c) Li, T.; Lohmann, F.; Famulok, M. Nat. Commun. 2014, 5, 4940. (5) (a) Mao, X. H.; Simon, A. J.; Pei, H.; Shi, J. Y.; Li, J.; Huang, Q.; Plaxco, K. W.; Fan, C. H. Chem. Sci. 2016, 7 (2), 1200−1204. (b) Song, P.; Li, M.; Shen, J.; Pei, H.; Chao, J.; Su, S.; Aldalbahi, A.; Wang, L.; Shi, J.; Song, S.; Wang, L.; Fan, C.; Zuo, X. Anal. Chem. 2016, 88 (16), 8043−9. (c) Adornetto, G.; Porchetta, A.; Palleschi, G.; Plaxco, K. W.; Ricci, F. Chem. Sci. 2015, 6 (7), 3692−3696. (6) Deng, M.; Zhang, D.; Zhou, Y.; Zhou, X. J. Am. Chem. Soc. 2008, 130 (39), 13095−102. (7) (a) Eichstaedt, K.; Jaramillo-Garcia, J.; Leigh, D. A.; Marcos, V.; Pisano, S.; Singleton, T. A. J. Am. Chem. Soc. 2017, 139 (27), 9376− 9381. (b) Kwan, C. S.; Chan, A. S. C.; Leung, K. C. F. Org. Lett. 2016, 18 (5), 976−979. (c) Rossetti, M.; Ranallo, S.; Idili, A.; Palleschi, G.; Porchetta, A.; Ricci, F. Chem. Sci. 2017, 8 (2), 914−920. (8) (a) Pelossof, G.; Tel-Vered, R.; Elbaz, J.; Willner, I. Anal. Chem. 2010, 82 (11), 4396−402. (b) Teller, C.; Shimron, S.; Willner, I. Anal. Chem. 2009, 81 (21), 9114−9. (c) Xiao, Y.; Pavlov, V.; Gill, R.; Bourenko, T.; Willner, I. ChemBioChem 2004, 5 (3), 374−9. (d) Zhu, X.; Gao, X.; Liu, Q.; Lin, Z.; Qiu, B.; Chen, G. Chem. Commun. (Cambridge, U. K.) 2011, 47 (26), 7437−9. (e) Eckhoff, G.; Codrea, V.; Ellington, A. D.; Chen, X. J. Syst. Chem. 2010, 1, 13. (f) Li, B.; Ellington, A. D.; Chen, X. Nucleic Acids Res. 2011, 39 (16), e110.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b08839. Experimental details (PDF) DNA rotaxane switching animation (MOV)

■

Communication

AUTHOR INFORMATION

Corresponding Authors

*[email protected] *[email protected] ORCID

Michael Famulok: 0000-0001-5878-6577 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS

We thank D. Ackermann and F. Lohmann for helpful discussions, D. Keppner for technical assistance, and the ERC (grant 267173), the Alexander von Humboldt Foundation and the Max-Planck Society for financial support. 16047

DOI: 10.1021/jacs.7b08839 J. Am. Chem. Soc. 2017, 139, 16044−16047