Fabrication of ordered metallic and magnetic heterostructured DNA—Nanoparticle hybrids

Available online at www.sciencedirect.com

Colloids and Surfaces B: Biointerfaces 63 (2008) 296–300

Short communication

Fabrication of ordered metallic and magnetic heterostructured DNA—Nanoparticle hybrids夽 Joseph M. Kinsella b , Albena Ivanisevic a,∗ a

Weldon School of Biomedical Engineering and Department of Chemistry, Purdue University, West Lafayette 47906, USA b Weldon School of Biomedical Engineering, Purdue University, West Lafayette 47907, USA Received 30 November 2007; received in revised form 7 December 2007; accepted 10 December 2007 Available online 23 December 2007

Abstract Here we provide a method based on enzymatically catalyzed reactions to cleave and ligate DNA molecules coated with nanoparticles to fabricate multi-component structures. This is done by simultaneously digesting two solutions of nanoparticle coated DNA, one with iron oxide particles the other gold particles, which yields short DNA fragments with complementary single stranded overhangs. When added together and re-attached using ligase enzymes multi-component nanoparticle coated structures are formed providing a novel method to fabricate complicated nanoparticle arrangements from the bottom up. We evaluated the fabrication by characterizing the samples with gel electrophoresis and magnetic force microscopy (MFM). The electrophoresis provides proof that the coated DNA molecules were digested with restriction enzymes and ligated by the T4 ligase enzymes. MFM experiments allow us to visualize the multi-component strands and analyze the magnetic versus metallic segments. © 2008 Elsevier B.V. All rights reserved. Keywords: Hybrid materials; Data storage; Magnetic nanoparticles; Scanning probe microscopy

1. Introduction Magnetic nanoparticles have been at the forefront of data storage research where next generation memory devices are anticipating the incorporation of nanoscale elements to increase data storage potential [1–3]. A variety of approaches, both conventional and unconventional, have been explored to provide applicable materials for this purpose [4–6]. One of the fundamental challenges faced is the organization of nanoparticles into sophisticated patterns where their nanoscale properties can be exploited [7,8]. This problem is not unique to data storage systems but extends to any potential magnetic, electronic, mechanical, or optical device incorporating nanoparticles were strict control over alignment and registration are required. This challenge becomes even more difficult when multiple types of nanoparticles are incorporated onto a single platform requiring precise control over alignment and registration relative to each other. Limited methods currently exist to address this problem and those that do are often material specific.

夽 ∗

This work was supported by NSF under CMMI-0727927. Corresponding author. E-mail address: [email protected] (A. Ivanisevic).

0927-7765/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfb.2007.12.004

One potential method to address this challenge is to utilize a bottom-up strategy using molecular guides to direct the positioning of nanoparticles. Using DNA as a molecular guide we can dictate the one-dimensional arrangement of positively charged magnetic nanoparticles along the length of the molecule using electrostatic interactions [9,10]. Other researchers have used DNA as a guide to organize metallic nanoparticles for electronic applications [11–13]. Typically molecular guides, or templates, are sacrificial elements that are only utilized to function as alignment tools. By using DNA as a guide we have chosen a molecule that has a variety of well-studied biochemical reactions which can be exploited. Another critical advantage in choosing DNA molecules as a template is the large aspect ratio provided by DNA. As our molecular guide we choose a DNA molecule that when stretched is over 20 ␮m in length and 2 nm in diameter [14,15]. Of particular interest to these studies are enzymatic reactions that site-specifically catalyze the fragmentation of DNA molecules and enzymatic reactions that re-connect the fragmented DNA strands. We have previously demonstrated that DNA coated with CoFe2 O4 magnetic nanoparticles can be cleaved with restriction enzymes and pieced back together with ligase enzymes [16]. Here we report using this system based on molecular biology to build one-dimensional arrangements of two different types of nanoparticles (gold and iron oxide) on

J.M. Kinsella, A. Ivanisevic / Colloids and Surfaces B: Biointerfaces 63 (2008) 296–300

297

Scheme 1. Experimental design used for creating an iron oxide and gold-heterostructured chain of particles using the restriction digest-ligation method. The products of the two reactions (a + b) are added together (c) T4 ligase is added to re-join the fragments yielding long DNA molecules containing site-specific coatings of both type of nanoparticle.

a single DNA strand. The specificity of the biomolecular template can provide an approach to do such with addressability of size specific segments. This is accomplished by first running a restriction digest using the BamHI restriction enzyme on two independent solutions of DNA each coated with a different type of nanoparticle. The results of this reaction yield one solution were DNA coated with iron oxide has been cleaved into 5-␮m sized fragments and another solution containing DNA coated with gold nanoparticles that has also been cleaved into five fragments. The fragments that are generated are self-complimentary, so mixing the two solutions together and adding the appropriate T4 ligase enzyme yields long segments of DNA that is coated with different types of nanoparticles at specific lengths along the DNA molecule. Scheme 1 shows a diagram depicting a typical experiment designed to form multi-component coated DNA with iron oxide and gold. We designed these experiments to be easily characterized by magnetic force microscopy (MFM) by ligating DNA coated with gold to DNA coated with Fe2 O3 . This combination of yields DNA that is coated in segments that are magnetic and will give a strong phase shift during MFM imaging as well as segments coated in gold which do not give any appreciable phase shift above 10 nm [9]. We also verified the ligation of the coated DNA fragments using gel electrophoresis. 2. Experimental 2.1. Nanoparticle synthesis

tom flask. The flask was heated to 200 ◦ C for a 30-min period. Nitrogen gas was added and the vessel was heated to reflux at 260 ◦ C. The solution was allowed to reflux for a 30-min period at which point the reaction vessel was removed from the heating source and cooled to room temperature. Methanol was added to the vessel once it reached room temperature to precipitate out the nanoparticles. The particles were precipitated from the solvent by adding 40 mL of methanol to the cooled solution. The particles were then centrifuged in acetone several times, dried under vacuum, and stored in water. Five nanometer poly-lysine capped Gold nanoparticles were purchased from Ted Pella. These particles were diluted 10-fold into a restriction digest buffer before use. 2.2. Nanoparticle coating of DNA molecules and biochemical reactions The coated DNA structures were formed by incubating 1 ␮g of lambda phage DNA (Promega, WI, USA) with an equivalent quantity of nanoparticles. The solution was then gently agitated on a vortex mixer for a period of greater than 1 h but less than 24 h. The samples were then treated with 10 units of BamHI restriction enzyme and allowed to react for a period of 4 h at room temperature. After this period, the samples were transferred to a hot plate at 65 ◦ C to denature the enzymes. The samples were then treated with a T4 ligase (Promega, WI, USA) solution in the appropriate buffer for 1 h at room temperature. 2.3. Magnetic force microscopy and gel electrophoresis

A 4-mmol solution of Iron(III) acetylacetonate (Fe(acac)3) (Sigma–Aldrich) dissolved in 40 mL of 2-pyrrolidinone (Sigma–Aldrich) was added to a nitrogen purged round bot-

A Multi-ModeTM Nanoscope IIIa from Digital Instruments, equipped with Nanoscope 5.12r5 software, was used to acquire

298

J.M. Kinsella, A. Ivanisevic / Colloids and Surfaces B: Biointerfaces 63 (2008) 296–300

Fig. 1. Height (A and C) and MFM phase (B and D) images of Fe2 O3 –Au heterostructured DNA. The MFM image in (B) was taken at 10-nm lift height and shows a uniform coating of the DNA as short-range forces convolute the sum of the imaging forces. In (D), the lift height was increased to 20 nm and segments of the DNA coated with gold nanoparticles lose contrast in the MFM scans. This is pointed out in the circled segment of (C) and (D). In the height scan (C), the coating appears continuous and no conclusion to particle type can be made. In the MFM scan (D) the segment inside the circle does not appear to be on the surface since it is a gold segment that does not show any shift in the phase from a lack of magnetic force. In (F), when the lift height is increased to 30 nm the iron oxide lacks sufficient magnetization to produce a phase shift.

the MFM data. MFM tips (Model NSC14/Co–Cr) with a spring constant of 5 N m−1 were purchased from Mikromasch. All the imaging experiments were performed under ambient conditions. The coated DNA was adsorbed onto silicon oxide substrates which were cleaned prior deposition with piranha and ultrasonication in water, ethanol, and methanol. Scan speeds of 0.5 Hz were used for MFM imaging. All MFM scans were performed at multiple lift heights until no phase shift was registered between the surface and tip. Gel electrophoresis experiments were performed using 0.8% agarose gels and run at 75 V with a constant current of 400 mA

for 4 h. Twenty microliter of solution containing each sample plus loading dye (Promega, WI, USA) were loaded into each lane. After completion the gels were stained in ethidium bromide, rinsed with water, and imaged in a gel dock with a UV camera. On each gel a 1 kb DNA ladder (Promega, WI, USA) was run to qualitatively compare migration distances. 3. Results MFM images of the products were collected at multiple lift heights to insure that the signal was occurring from magnetic

J.M. Kinsella, A. Ivanisevic / Colloids and Surfaces B: Biointerfaces 63 (2008) 296–300

fields emanating from the Fe2 O3 nanoparticles and not surface features that are non-magnetic. In Fig. 1, we present height images (left column) and corresponding MFM images (right column) of the Fe2 O3 –Au heterostructured DNA at (A and B) 10 nm lift height, (C and D) 20 nm lift height, and (E and F) 30 nm lift height. In all height images (A, C, and E) the coated DNA strands appear the same since the MFM tip is in close contact with the surface and short-range forces dominate the image formation. In the MFM scans, when the tip-sample distance is increased the segments coated in gold show no contrast since they are not magnetic and appear as gaps along the DNA. This is shown in (D) a 20 nm lift height scan collected simultaneously with the height scan in (C). The circled area shows a section of a DNA strand coated with gold that appears to have zero con-

299

trast at this lift height. In the height scan, this section of the DNA appears continuously coated with nanoparticles though no information on the type of nanoparticles can be gained since the size of both particle types is below that of tip convolution effects. In the MFM scan show in (E) the distance of the tip is to far from the sample to interact with the Fe2 O3 nanoparticles and little phase shift occurs on any section of the coated DNA. We also evaluated the effectiveness of the fabrication procedure using gel electrophoresis (Fig. 2). This allowed us to run several reactions in parallel in solution and perform control experiments with DNA that was not coated by any nanoparticles. The controls that were run include uncoated DNA which was not treated to any restriction digestion reactions or ligation reactions and uncoated DNA that was cleaved with BamHI and subsequently ligated with T4 ligase. These controls (Lanes 1 and 2) provide a means to monitor the migration distance in the gels and provide a comparison to coated DNA that underwent enzymatic reactions. In the controls, we noticed that the uncoated DNA that did not receive any enzymatic treatment migrated a short distance further in a 0.8% agarose gel (Lane 1) than uncoated DNA that was digested and ligated (Lane 2). This occurs since the ligation of self-complimentary DNA fragments can lead to a variety of different combinations of fragments some of which will have a slightly higher number of base pairs than the original strand. In Lane 3, we ligated DNA fragments that were coated with Fe2 O3 and DNA fragments coated with gold nanoparticles. This solution contains the multi-component DNA strands that underwent the same treatment as samples used to generate the MFM data. In the gel experiments, the ligated solution yielded migration distances that agree with the control experiments indicating that the ligation was run to completion during the time allotted and the resulting DNA molecules were >16 ␮m in length. In Lanes 4 and 5, iron oxide and goldcoated DNA was ligated with uncoated DNA to insure that the particular particle types would not interfere with the ligation reactions. Neither iron oxide nor gold showed any difference in the migration distance when ligated to uncoated DNA compared to control experiments or ligation experiments were both solutions of digested DNA were coated with a type of nanoparticle. 4. Conclusions

Fig. 2. Gel electrophoresis experiments of multi-component coated DNA. Lane 1 is bare DNA that was not treated with restriction enzymes or T4 ligase and serves as a reference for migration distance. Lane 2 shows uncoated DNA that was digested and ligated and serves as a secondary migration distance reference. Lane 3 shows the products of iron oxide coated DNA ligated to gold-coated DNA. Lane 4 shows the products of iron oxide coated DNA ligated with uncoated DNA. Lane 5 shows gold-coated DNA ligated with uncoated DNA.

The intention of the experiments was to determine whether we could utilize enzymatic methods for cleaving DNA and ligating self-complimentary fragments to build DNA strands coated with multiple materials. The experiments that we performed indicate this is a viable means to do such. We demonstrated this by building DNA strands that were coated by two types of nanoparticles, gold and iron oxide, and ligating the mixed products of BamHI restriction digestions to generate long heterostructured nanoparticle chains. MFM experiments verified this by allowing us to map areas of the DNA that are magnetic versus those that are not magnetic with nanometer scale resolution. Gel electrophoresis verified the ubiquity of these enzymatic reactions by providing evidence that any configuration of coated or uncoated DNA can be first digested with a particular restric-

300

J.M. Kinsella, A. Ivanisevic / Colloids and Surfaces B: Biointerfaces 63 (2008) 296–300

tion enzyme and then combined and ligated to form DNA strands of equal length to the starting DNA molecules. Biochemical reactions have not been explored on coated DNA to create such two component structures prior to this demonstration. From what we have learned with these experiments it should be possible to extrapolate these reactions to systems containing more than two types of nanoparticle coatings and build more sophisticated one-dimensional alignment schemes. These enzymatic methods provide a means to obtain site-specific control over which nanoparticle resides along the ligated multi-component structures if dual restriction digests are utilized. This could lead to control over not only spacing of the type of nanoparticle coated DNA along the ligated strand, but also control over the registration of the types of nanoparticle coated DNA. Given proper experimental design and control this method could be used to generate sophisticated multi-component magnetic, electrical, or optical devices with nanoscale dimensions. References [1] J.M. Daughton, Magnetoresistive memory technology, Thin Solid Films 216 (1992) 162–168. [2] K. O’Grady, H. Laidler, The limits to magnetic recording—media considerations, J. Magn. Magn. Mater. 200 (1999) 616–633. [3] C. Ross, Patterned magnetic recording media, Ann. Rev. Mater. Res. 31 (2001) 203–235. [4] L. Fu, X.G. Liu, Y. Zhang, VP. Dravid, C.A. Mirkin, Nanopatterning of “hard” magnetic nanostructures via dip-pen nanolithography and a solbased ink, Nano Lett. 3 (2003) 757–760.

[5] X.G. Liu, L. Fu, S.H. Hong, V.P. Dravid, C.A. Mirkin, Arrays of magnetic nanoparticles patterned via “dip-pen” nanolithography, Adv. Mater. 14 (2002) 231–234. [6] S. Palacin, P.C. Hidber, J.P. Bourgoin, C. Miramond, C. Fermon, G.M. Whitesides, Patterning with magnetic materials at the micron scale, Chem. Mater. 8 (1996) 1316–1325. [7] J. Chang, V.L. Mironov, B.A. Gribkov, A.A. Fraerman, S.A. Gusev, S.N. Vdovichev, Magnetic state control of ferromagnetic nanodots by magnetic force microscopy probe, J. Appl. Phys. 100 (2006). [8] C.R. Vestal, Q. Song, Z.J. Zhang, Effects of interparticle interactions upon the magnetic properties of CoFe2 O4 and MnFe2 O4 nanocrystals, J. Phys. Chem. B 108 (2004) 18222–18227. [9] J.M. Kinsella, A. Ivanisevic, DNA-templated magnetic nanowires with different compositions: fabrication and analysis, Lang 23 (2007) 3886– 3890. [10] D. Nyamjav, A. Ivanisevic, Templates for DNA-templated Fe3 O4 nanoparticles, Biomaterials 26 (2005) 2749–2757. [11] G. Braun, K. Inagaki, R.A. Estabrook, D.K. Wood, E. Levy, A.N. Cleland, G.F. Strouse, N.O. Reich, Gold nanoparticle decoration of DNA on silicon, Lang 21 (2005) 10699–10701. [12] W.U. Dittmer, F.C. Simmel, Chains of semiconductor nanoparticles templated on DNA, Appl. Phys. Lett. 85 (2004) 633–635. [13] J. Lund, J.C. Dong, Z.X. Deng, C.D. Mao, B.A. Parviz, Electrical conduction in 7 nm wires constructed on lambda-DNA, Nanotechnology 17 (2006) 2752–2757. [14] A. Bensimon, A.J. Simon, A. Chiffaudel, V. Croquette, F. Heslot, D. Bensimon, Alignment and sensitive detection of DNA by a moving interface, Science 265 (1994) 2096–2098. [15] D. Bensimon, A.J. Simon, V. Croquette, A. Bensimon, Stretching DNA with a receding meniscus: experiments and methods, Phys. Rev. Lett. 74 (1995) 4754–4757. [16] J.M. Kinsella, M.V. Shalaev, A. Ivanisevic, Ligation of nanoparticle coated DNA cleaved with restriction enzymes, Chem. Mater. 19 (2007) 3586–3588.