communications Gold–DNA assemblies DOI: 10.1002/smll.200600494
Enzymatic Disassembly of DNA–Gold Nanostructures** Antonios G. Kanaras, Zhenxin Wang, Mathias Brust, Richard Cosstick, and Andrew D. Bates* The current success of interdisciplinary research in nanoscience is generating great expectations for the emergence of new technologies. It is widely believed that significant advances in electronics, information technology, sensor development, catalysis, and biomedical sciences will arise from gaining precise control over the manipulation of nanometersized objects. The interactions that govern the behavior of matter on this scale are predominantly chemical in nature. Therefore, the development of new chemical tools for the controlled assembly and manipulation of nanostructures is of great interest. An important step has been the development of so-called programmed assembly, where connectivity between building blocks is predetermined by specific recognition between biomolecular or synthetic connectors.[1–5] For instance, excellent specificity and control over the binding process can be achieved by exploiting DNA base-pair recognition for the attachment of DNA-modified nanoparticles to each other by hybridization with complementary linker strands.[6–7] Using this approach, originally developed by Alivisatos and Mirkin and co-workers, it is possible to design particles that will selectively bind to other particles of a particular type or to designated sites on surfaces.[8–20] There are well-established biomolecular methods for DNA manipulation, based on the use of DNA-processing enzymes (restriction endonucleases and ligases), to cleave and rejoin double-stranded DNA with impressive site specificity.[21–22] Using a combination of these enzymes and thiol-modified DNA attached to gold nanoparticles, we have recently dem-
[*] Dr. A. D. Bates School of Biological Sciences, The University of Liverpool Liverpool L69 7ZB (UK) Fax: (+ 44)151-795-4410 E-mail:
[email protected] Dr. A. G. Kanaras, Prof. Z. Wang, Prof. M. Brust, Dr. R. Cosstick Centre for Nanoscale Science, Department of Chemistry The University of Liverpool, Liverpool L69 7ZD (UK) Prof. Z. Wang Current address: State Key Laboratory of Electro-analytical Chemistry Changchun Institute of Applied Chemistry Chinese Academy of Science, 5625 Renmin Street Changchun, 130022 (P.R. China) [**] Financial support by the European Union (BIOSCOPE) is gratefully acknowledged. Research was partially supported by the University of Liverpool. Z.W. acknowledges support from the BBSRCfunded Liverpool Centre for Bioarray Innovation. Supporting information for this article is available on the WWW under http://www.small-journal.com or from the author.
590
onstrated hierarchical and temporal control of assembly of DNA-modified gold nanoparticles. In particular, restriction endonucleases can be used to reveal reactive DNA ends on DNA-modified particles, which may then be converted to permanent links with high specificity between particles (and between particles and surfaces) using DNA ligase.[23–25] In this Communication we report our recent progress in using DNA-processing enzymes for the specific cleavage of preassembled DNA-modified gold nanostructures. Yun et al.[26] have suggested this possibility using dimers of DNA-modified gold nanoparticles, based on a statistical analysis of transmission electron microscopy (TEM) images before and after the enzymatic reaction step. We now demonstrate, using a number of complementary techniques, that aggregates of gold nanoparticles linked to each other by double-stranded DNA can be efficiently cleaved and redispersed by restriction enzymes. We synthesized assemblies of DNA-modified gold nanoparticles following the protocols reported by Mirkin et al.[27] Initially, a three-strand DNA system was prepared from two batches of 15 nm gold particles derivatized with distinct single-stranded DNAs and low concentrations of a complementary ssDNA as linker (Figure 1). The DNA between the particles was designed to contain a recognition site for the restriction endonuclease EcoRI. The mixture was then incubated with the enzyme. The enzymatic reaction resulted in a redispersion of the cloudy aggregates to form a clear solution (see Supporting Information). TEM images (Figure 2 A) showed that three-dimensional aggregations of particles prior to the enzymatic cleavage were indeed dispersed after reaction with EcoRI. The small two-dimensional aggregations present after the enzymatic cleavage (Figure 2 A) may be due to weak association between the four-nucleotide (4-nt) :cohesive ends; generated on half of the resulting single particles (Figure 1 B).[23] Based on the stoichiometry used, only around 10 % of the 150 DNA strands per particle[24] were joined with linker strands, so after cleavage, one half of the particles would have about fifteen 4-nt cohesive ends each. The other half of the particles would have 2-nt overhangs, which will form only very unstable associations. Further evidence for the enzymatic cleavage of the DNA–particle assemblies comes from the change in the optical properties of the system. In a spot test, where a drop of the solution was deposited on a reverse-phase silica plate,[9] the color of the spot changed from blue to red after the enzymatic reaction (Figure 2 B). There was a corresponding shift in the visible absorption maximum from 534 nm to 524 nm (Figure 2 C). That this color change was due to DNA cleavage is confirmed by varying the temperature. The spot test performed after the enzymatic cleavage remained red at any temperature (DNA cleaved) and the absorbance maximum was unchanged. For uncleaved assemblies, the spot test changed from blue to red and the absorbance maximum shifted only at elevated temperatures (when the DNA undergoes melting); these changes were reversed on cooling to room temperature (DNA rehybridized). We also analyzed DNA cleavage by polyacrylamide gel electrophoresis (PAGE). After release of the DNA from the particles using an excess of dithiothreitol (DTT),[23] the
A 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2007, 3, No. 4, 590 – 594
the gel. This uncleaved linker could contribute to the residual small assemblies visible in the TEM (Figure 2 A). We compared the cleavage of nanoparticle assemblies constructed using a three-strand linker with those using a simpler twostrand system, in which two samples of particles derivatized with complementary DNAs are mixed and allowed to hybridize (see Supporting Information for sequences). In this case, there is no control over the number of links between particles, which in the threestrand system is afforded by varying the concentration of the linker strand. The separation of these particles by EcoRI restriction enzyme is much less efficient, with fairly large assemblies visible in the TEM after treatment (Figure 4 A). This effect is confirmed by the optical properties (Figure 4 B, C). InFigure 1. A) Schematic view of 15-nm gold nanoparticle assemblies joined by a three-strand system. itially, the aggregates are DNA is in grey, with the EcoRI recognition site in black, and positions of cleavage indicated by arrows. In more thermally stable, with reality, only 10 % of the single-strand DNA ligands are linked by a third strand (based on the stoichiomethe spot tests indicating try of the linking strand). B) The sequences of the three (1, 2, 3) DNA strands are shown, with the EcoRI closely spaced particles up to site in black. After cleavage by the enzyme at the arrowed sites, a 2-nt fragment (AA) is lost, leaving par85 8C. After treatment, therticles with 4-nt overhanging (cohesive) ends, and particles with 2-nt cohesive ends. mal stability is reduced, but the particles are clearly still aggregated at 25 8C, with an absorption maximum (530 nm) in between that of fully disDNA was analyzed by electrophoresis (Figure 3). Before persed and fully aggregated particles. EcoRI treatment (Lane 1), a band of around 45 base pairs This difference in behavior probably reflects the lack of (bp) is visible, corresponding to the full-length doublecontrol over particle linking in the two-strand system. The stranded linker. When DNA was recovered from particles number of links will be much larger than in the three-strand after EcoRI treatment (Lane 2), this is replaced by a broad case, where the linking stoichiometry is limited, leading to band of approximately 25 bp consistent with a mixture of increased stability. It may be that enzyme action is limited the two cleaved DNAs (Figure 1 B). It is highly likely that by steric inaccessibility of the cleavage sites in a dense netcleavage occurred specifically at the EcoRI recognition site. work of crosslinks, or cleaved crosslinks may be replaced by DNA cleavage by restriction enzymes is known to be highly newly hybridized single strands; hence dispersion of the parspecific for the recognition sequence.[22] We have shown preticles by cleavage is much less efficient. viously that this specificity is maintained in the context of In conclusion, we have demonstrated the efficient enzyDNA complexed with nanoparticles, in experiments where matic cleavage of DNA-modified gold networks by a DNA cohesive ends are revealed on particles by restriction number of complementary techniques. However, the effienzyme action, and can then be covalently ligated to specific ciency of the process is dependent on the design of the complementary sites to form links between particles.[23, 25] linker system, with the more controllable assembly afforded These earlier results also confirm that the DNA ligands are by the three-strand linking system (Figure 1) being much not stripped from the gold particles by the enzyme treatmore effectively manipulated by a restriction enzyme. The ment itself, and therefore the cleavage reaction reported ability to manipulate DNA enzymatically in these situations here must be taking place in the intact nanoparticle assemis an important proof-of-principle in the development of hibly. A small fraction of uncleaved DNA is apparent on the erarchical chemical methods of nanoscale assembly. Such gel (Figure 3, Lane 2), estimated at < 5 % from analysis of small 2007, 3, No. 4, 590 – 594
A 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com
591
communications
Figure 2. TEM and optical properties of three-strand nanoparticle assemblies before (left) and after (right) treatment with EcoRI. A) TEM: Small residual assemblies after treatment may be due to weak interactions between 4-nt cohesive ends, or to incomplete DNA cleavage. B) Spot tests. C) Absorbance spectra: Before cleavage, assemblies show properties characteristic of closely spaced particles (blue spot test, 534-nm broad absorption maximum) at room temperature. The particles can be reversibly separated at high temperature by dehybridization. After cleavage, particles are separated at all temperatures (red spot test, 524-nm sharp maximum).
Preparation of gold nanoparticles: Gold nanoparticles (15 nm; stabilized electrostatically with citrate anions) were prepared by reduction of tetrachloroaurate as reported previously.[28, 29] DNA-modification of gold nanoparticles: Citrate-stabilized gold nanoparticles were derivatized with thiolated oligonucleotides as previously described.[23] Three-strand system: Two different batches of DNA-modified gold particles were prepared and complementary ssDNA (1 mL, 50 mm) was added to a solution containing both batches of particles (200 mL, each at 30 nm) in hybridization buffer (sodium phosphate (10 mm, pH 7); NaCl (0.3 m)). During the next 5 h a gradual aggregation took place, and the color of the solution changed from red to blue. Subsequently, the solution was heated to 85 8C and the color changed back to red. Two-strand system: Two batches of DNA/gold particles with complementary DNA ligands (see Supporting Informa-
cleavage might be designed to deprotect sites for further synthetic assembly by DNA ligation[23, 25] or could find application in the controlled release of proteins or other biomolecules in solution.
Experimental Section Hydrogen tetrachloroaurate trihydrate (HAuCl4·3H2O), sodium borohydride, and tetraoctylammonium bromide were purchased from Sigma. Solvents, sodium citrate, and other salts were supplied from BDH. For all experiments Milli-Q water (18.1 MW) was used, purified with an ultrapure water system Milli-Q Plus 185 (Millipore purification pack). 0.45 mm pore size Millex-HA microfilters were purchased from Millipore, oligonucleotides were purchased from Sigma-Genosys and EcoRI was purchased from Roche.
592
www.small-journal.com
Figure 3. Nondenaturing polyacrylamide gel electrophoresis of DNA. Lanes 1 and 2, respectively, show DNA recovered from particles before and after the particle assemblies had been treated with EcoRI. The positions of the different DNA species are indicated by pictograms from Figure 1. Lane M contains double-stranded DNA standards of the indicated sizes in base pairs. The gel is 20 % (w/v) polyacrylamide (19:1 acrylamide:bisacrylamide) in 90 mm Trisborate, 1 mm EDTA, pH 8.0, stained with ethidium bromide.
A 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2007, 3, No. 4, 590 – 594
400-mesh grid. All samples were examined in a JEOL 2000 EX TEM operating at 200 kV. Gel electrophoresis: Gel electrophoresis of DNA released from gold particles with DTT (as previously described[23]) was carried out on 20 % (w/v) polyacrylamide (19:1 acrylamide:bisacrylamide) in TBE (Trisborate (90 mm, pH 8.0); EDTA (1 mm)) and the bands were visualized by staining with ethidium bromide.
Keywords: DNA · gold · nanoparticles · nanostructures · restriction enzymes
[1] J. J. Storhoff, C. A. Mirkin, Chem. Rev. 1999, 99, 1849 – 1862. [2] C. M. Niemeyer, Angew. Chem. 2001, 113, 4254 – 4287; Angew. Chem. Int. Ed. 2001, 40, 4128 – 4158. [3] W. J. Parak, D. Gerion, T. Pellegrino, D. Zanchet, C. MiFigure 4. TEM and optical properties of nanoparticle assemblies with two-strand linkers before (left) and cheel, S. C. Williams, R. Bouafter (right) treatment with EcoRI. A) TEM: Larger residual aggregates than in Figure 2 are visible after dreau, M. A. Le Gros, C. A. enzyme cleavage. B) Spot tests indicate that, unlike in the three-strand system (Figure 2), the particles Larabell, A. P. Alivisatos, remain closely spaced after enzyme treatment (blue spot test), albeit with a lower thermal stability than Nanotechnology 2003, 14, before cleavage. C) Absorbance spectra: At 25 8C, the cleaved assembly has an absorption maximum 15 – 27. (530 nm) intermediate between those of the fully formed (540 nm) and fully dispersed (524 nm) assem[4] E. Katz, I. Willner, Angew. blies at low and high temperatures, respectively, before cleavage. Chem. 2004, 116, 6166 – 6235; Angew. Chem. Int. Ed. 2004, 43, 6042 – 6108. tion) were mixed (150 mL, 15 nm of each) and left to stand for [5] T. Pellegrino, S. Kudera, T. 5 h in hybridization buffer. Liedl, A. M. Javier, L. Manna, W. J. Parak, Small 2005, 1, 48 – 63. [6] C. A. Mirkin, R. L. Letsinger, R. C. Mucic, J. J. Storhoff, Nature Enzymatic cleavage of DNA–gold aggregates: a) Three-strand 1996, 382, 607 – 609. system: Aggregated DNA–gold colloids (7.5 nm final particle con[7] A. P. Alivisatos, K. P. Johnsson, X. Peng, T. E. Wilson, C. J. centration) were incubated overnight at 37 8C in the presence or Loweth, M. P. Bruchez Jr, P. G. Schultz, Nature 1996, 382, 609 – absence of EcoRI (200 units) in 400 mL Tris-HCl (100 mm, 611. pH 7.5), NaCl (50 mm), and MgCl2 (10 mm). EDTA was then [8] R. Elghanian, J. J. Storhoff, R. C. Mucic, R. L. Letsinger, C. A. added to 20 mm and the solution left to stand for 1 h. The soluMirkin, Science 1997, 277, 1078 – 1081. tion was centrifuged twice (Sigma 1–13 microcentrifuge, [9] J. J. Storhoff, R. Elghanian, R. C. Mucic, C. A. Mirkin, R. L. Letsinger, J. Am. Chem. Soc. 1998, 120, 1959 – 1964. 13,000 rpm, 25 min) and the pellet resuspended in hybridization [10] C. L. Loweth, W. B. Caldwell, X. Peng, A. P. Alivisatos, P. G. buffer. b) Two-strand system: As above, with particles (2.5 nm Schultz, Angew. Chem. 1999, 111, 1925 – 1929; Angew. Chem. final particle concentration) and EcoRI (300 units). Int. Ed. 1999, 38, 1808 – 1812. UV/Vis spectra: UV/Vis spectra were recorded on a Gen[11] S. Mann, S. Shenton, M. Li, S. Connolly, D. Fitzmaurice, Adv. esys 10-series spectrometer from ThermoSpectronic. Mater. 2000, 12, 147 – 150. Spot tests: Spot tests of the aggregation state of the parti[12] J. J. Storhoff, A. A. Lazarides, R. C. Mucic, C. A. Mirkin, R. L. Letsinger, G. C. Schatz, J. Am. Chem. Soc. 2000, 122, 4640 – 4650. cles were carried out by spotting a suspension of the particles [13] T. A. Taton, C. A. Mirkin, R. L. Letsinger, Science 2000, 289, on a reverse-phase silica TLC plate as described previously.[8] 1757 – 1760. TEM: Specimens were prepared by deposition of a drop of
an aqueous solution of the particles on a carbon-coated copper small 2007, 3, No. 4, 590 – 594
A 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com
593
communications [14] D. Zanchet, C. M. Micheel, W. J. Parak, D. Gerion, A. P. Alivisatos, Nano Lett. 2001, 1, 32 – 35. [15] E. Dujardin, S. Mann, Adv. Mater. 2002, 14, 775 – 788. [16] C. M. Niemeyer, B. Ceyhan, P. Hazarika, Angew. Chem. 2003, 115, 5944 – 5948; Angew. Chem. Int. Ed. 2003, 42, 5766 – 5770. [17] P. Hazarika, B. Ceyhan, C. M. Niemeyer, Angew. Chem. 2004, 116, 6631 – 6633; Angew. Chem. Int. Ed. 2004, 43, 6469 – 6471. [18] A. H. Fu, C. M. Micheel, J. Cha, H. Chang, H. Yang, A. P. Alivisatos, J. Am. Chem. Soc. 2004, 126, 10 832 – 10 833. [19] S. A. Claridge, S. L. Goh, J. M. J. Frechet, S. C. Williams, C. M. Micheel, A. P. Alivisatos, Chem. Mater. 2005, 17, 1628 – 1635. [20] J. S. Lee, S. I. Stoeva, C. A. Mirkin, J. Am. Chem. Soc. 2006, 128, 8899 – 8903. [21] H. O. Smith, D. Nathans, J. Mol. Biol. 1973, 81, 419 – 423. [22] A. Pingoud, A. Jeltsch, Nucleic Acids Res. 2001, 29, 3705 – 3727.
594
www.small-journal.com
[23] A. G. Kanaras, Z. Wang, A. D. Bates, R. Cosstick, M. Brust, Angew. Chem. 2003, 115, 201 – 204; Angew. Chem. Int. Ed. 2003, 42,191 – 194. [24] Z. Wang, A. G. Kanaras, A. D. Bates, R. Cosstick, M. Brust, J. Mater. Chem. 2004, 14, 578 – 580. [25] A. G. Kanaras, Z. Wang, I. Hussain, M. Brust, R. Cosstick, A. D. Bates, Small 2007, 3, 67 – 70. [26] C. S. Yun, G. A. Khitrov, D. E. Vergona, N. O. Reich, G. F. Strouse, J. Am. Chem. Soc. 2002, 124, 7644 – 7645. [27] L. M. Demers, C. A. Mirkin, R. C. Mucic, R. A. Reynolds III, R. L. Letsinger, G. Viswanadham, Anal. Chem. 2000, 72, 5535 – 5541. [28] J. Turkevich, P. S. Stevenson, J. Hillier, Discussions Faraday Soc. 1951, 11, 55 – 75. [29] G. Frens, Nature Phys. Sci. 1973, 241, 20 – 22.
A 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Received: September 13, 2006 Revised: January 18, 2007 Published online on February 22, 2007
small 2007, 3, No. 4, 590 – 594
