Self-Organized Inorganic Nanoparticle Arrays on Protein Lattices

Department of Physics, Meiji UniVersity, Higasimita, Tama-ku,. Kawasaki, 214-8571, Japan. Received March 23, 2005. ABSTRACT. Cavities formed by protei...
1 downloads 0 Views 308KB Size
NANO LETTERS

Self-Organized Inorganic Nanoparticle Arrays on Protein Lattices

2005 Vol. 5, No. 5 991-993

Mitsuhiro Okuda,† Yusaku Kobayashi, Kazuomi Suzuki, Katsuhisa Sonoda, Tetsuro Kondoh, Akitoshi Wagawa, Akira Kondo, and Hideyuki Yoshimura* Department of Physics, Meiji UniVersity, Higasimita, Tama-ku, Kawasaki, 214-8571, Japan Received March 23, 2005

ABSTRACT Cavities formed by proteins have been utilized as the reaction chamber for the fabrication of a range of inorganic nanoparticles,1-3 providing control of the size of particles by limiting growth and preventing agglomeration. In crystal form, proteins construct molecular arrays that can provide regularly arranged sites for nanoparticles. Here we report the fabrication of nanometric iron and indium particles using ferritin, an iron-storage protein. The indium nanoparticles thus formed have uniform spherical shape with diameter of 6.6 ± 0.5 nm, while the iron nanoparticles are somewhat irregular in shape (5.8 ± 1.0 nm). Regular two-dimensional arrays of these nanoparticles are successfully produced by crystallizing ferritin molecules on a water−air interface using the denatured protein film method.4 The lattice constant of these nanoparticle arrays is 13 nm with hexagonal packing, and arrays of more than 1 µm2 in area can be obtained by transfer onto silicon wafer.

Nanometric inorganic particles have a range of applications and are of particular utility when arranged in a regular lattice. A number of approaches for the fabrication of nanoinorganic materials have been reported based on physical or chemical methods, yet it remains difficult to array the particles at regular lattice points due to tendency for agglomeration. Surfactants have been used to coat nanoparticles as means of preventing agglomeration, but this leads to a loss of interaction specificity between nanoparticles and renders it impractical to realize regular arrays similar to those in the crystal state. The use of protein cages instead of surfactants, however, should impose monodispersity and specificity as inherent characteristics of proteins. The protein ferritin is spherical with a diameter of 12 nm and internal cavity of 7 nm, in which the protein forms a ferric hydroxide core through its main function of oxidizing ferric ion.5,6 Ferritin is also known to form cores of other metal compounds such as manganese,1 cobalt,7 nickel,8 and chromium.8 It has been shown that apoferritin (core-less ferritin) can be crystallized in two-dimensions by spreading an apoferritin solution on a water-air interface.4 If ferritin could be used in this manner instead of apoferritin, subsequent crystallization would form an array of metal cores positioned at the lattice points. As the crystal form can be controlled by modification of amino acids on the surface of apoferritin,9 various arrangements of nanoparticles should therefore be possible. Here we report the formation of iron and indium oxide cores in apoferritin * Corresponding author. E-mail: [email protected], Tel: +81-44934-7439, Fax: +81-44-900-0421. † Present address: Advanced Technology Research Laboratories, Panasonic, 3-4 Hikaridai, Seika, Kyoto, 619-0237, Japan. 10.1021/nl050556q CCC: $30.25 Published on Web 04/26/2005

© 2005 American Chemical Society

cavities and the realization of regular two-dimensional arrays of nanoparticles by crystallizing proteins on a water-air interface. Recombinant apoferritin was cloned from horse liver L-apoferritin10 and prepared as reported in the literature.8 Although the reconstitution of iron cores in horse spleen apoferritin has been reported by many authors,5,6 this study demonstrates improved conditions for iron uptake in order to achieve perfect core formation without remnant coreless apoferritin. The iron cores were formed in the apoferritin cavities by incubating 0.3 mg/mL of recombinant apoferritin in 100 mM HEPES (pH 7.5) and 3 mM ammonium iron sulfate at 4 °C. A brown precipitate was observed to form in the protein-free experiment, whereas no such precipitation occurred in the solution containing the protein after 24 h, indicating that the iron oxide had formed in the cavity of the recombinant apoferritin. After centrifugation (4000 g × 30 min) to remove the bulk salt precipitate, the supernatant was passed through a 450 nm filter and condensed by a centrifugal filter (Centriprep 50, Amicon). The ferritin was purified further by chromatography (Sephadex G-25 and Sephacryl S-300 columns) to remove free iron ions and aggregations of ferritin molecules. Finally, the ferritin was concentrated again by centrifugal filtration (Centriprep 50) to 3.0 mg/mL. Core formation was checked by transmission electron microscopy (TEM; JEM1200EX, JEOL) using aurothioglucose negative staining to distinguish apoferritin from ferritin.8 About 94% of the recombinant apoferritin had formed iron cores by this procedure. An unstained TEM image of the cores is shown in Figure 1a. Energy-dispersive

Figure 1. Unstained TEM images of (a) iron and (b) indium cores.

Figure 3. Unstained TEM image of iron core array formed by transferring two-dimensional crystal film of Fe-ferritin formed on a water-air interface to a holey carbon grid.

Figure 2. Size distribution of cores of Fe-ferritin and In-ferritin.

X-ray analysis (EDX) by TEM at core area showed peaks of iron (KR: 6.4 keV, Kβ: 7.1 keV, and L: 0.71 keV) and oxygen (KR: 0.53 keV). The iron cores exhibited some variation in shape and density. An average diameter which estimated from the area of individual particle was 5.8 ( 1.0 nm (Figure 2). Two-dimensional crystallization was performed according to a method reported previously4 with minor modification of the subphase solution. A circular Teflon trough (depth 5 mm, diameter 15 mm) was filled with 500 µL of the subphase solution (10 mM MES pH 5.8, 2% glucose, 10 mM CdSO4), followed by the injection of 1 µL of purified iron ferritin solution (3.0 mg/mL) underneath the subphase. The Feferritin (ferritin with iron core) rose to the water-air interface due to buoyancy in glucose. A small fraction of the ferritin units denatured at the water-air interface to form an unfolded polypeptide film. Intact ferritin molecules adsorbed on the denatured film underwent self-arrangement to form a twodimensional crystal through cadmium salt bridging. The twodimensional crystals were transferred to TEM holey carbon grids by a horizontal transfer method and subsequently reinforced by carbon coating.4 The TEM image of the iron array is shown in Figure 3. A regular array of iron cores with area of more than 1 µm2 in area was obtained, although the array contains a number of defects. These defects were not observed when the crystal was negatively stained by uranyl acetate, indicating that the defect sites are occupied by coreless apoferritin. The crystals were also transferred onto a silicon wafer (5 mm × 5 mm × 0.5 mm) covered with a 100 nm-thick silicon oxide layer. To accomplish this, the subphase solution was drained slowly 10 min after spreading the ferritin, and the solution remaining on the 992

silicon wafer was removed by centrifugation (5000 g × 10 min). Scanning electron microscopy (SEM; S-5200, Hitachi) of the two-dimensional crystals on the silicon wafer revealed arrays similar to those shown in Figure 1. The indium cores were similarly formed in the cavity of the recombinant apoferritin by incubation of apoferritin solution with indium sulfate at room temperature for 24 h. The solution pH was adjusted to around 2.8 with 40 mM hydrochloric acid, 200 mM monobasic sodium phosphate, and 4 mM ammonium. The concentration of recombinant apoferritin was 0.1 mg/mL in the total sample (150 mL). Indium sulfate solution (20 mM) was added gradually to achieve the final concentration of 1 mM. A white precipitate was observed to form in the protein-free experiment, and the reaction in substantially less white precipitate was formed in the protein solution after 24 h. The In-ferritin (ferritin with indium core) was subsequently purified and concentrated by the same procedures as for Fe-ferritin. TEM observations showed that almost 100% of the apoferritin contained In cores. An unstained TEM image of the cores is shown in Figure 1b. The cores were confirmed to be spherical by the circular projection in any orientation. The size and density of particles were uniform with an average diameter of 6.6 ( 0.5 nm. EDX of the core revealed specific peaks due to indium (LR:3.30 keV, Lβ1: 3.48 keV, Lβ2: 3.71 keV), phosphate (KR: 2.0 keV), and oxygen (KR: 0.53 keV). A two-dimensional crystal film of the In-ferritin was made by the same procedure as for Fe-ferritin. Indium dots in a hexagonal packing arrangement were successfully obtained on both holey carbon mesh and silicon wafers (Figure 4), with an area of up to 4 µm2. The indium dot array contained fewer defects than the iron dot array, due primarily to the more quantitative uptake of In. Several mechanisms for iron incorporation into apoferritin cavities have been proposed.11 The common features of all mechanisms, however, are the flow of Fe2+ through the negatively charged ion channel around the 3-fold symmetry axis of the apoferritin molecule, followed by the oxidation of Fe2+ to Fe3+ at oxidation sites on the inner surface, and finally growth of the iron core at the nucleation site. The Nano Lett., Vol. 5, No. 5, 2005

In summary, iron and indium nanoparticle arrays were successfully fabricated on carbon film and silicon wafer using a protein crystal lattice. After removal of the protein template by heat treatment12 or chemical treatment after adsorption onto a solid surface, these nanodot arrays may be applicable in semiconductor fabrication as a mask for X-ray lithography or particle etching and as an element of the nanotransistor array itself. The indium oxide nanoarray in particular has potential for use in a wide range of electronic devices.

Figure 4. SEM image (secondary electron image) of indium core array.

oxidation sites do not appear to be crucial for this process, as L-apoferritin lacks such sites and was shown in this study to form iron cores. The oxidation or hydration processes that induce sedimentation of metal salts (e.g., as manganese, cobalt, nickel, or chromium) are also considered to occur in the apoferritin cavity as a regular inorganic reaction of Fe in solution. Sedimentation of these metal salts occurs above pH 7, where acidic amino acids such as glutamic acid and aspartic acid dissociate with negative charge. These charged amino acids are thought to play an important role in ion transport and nucleation in apoferritin. As apoferritin consists of 24 isomers, there exist several identical ion channels and nucleation sites. Iron core formation can therefore initiate from several points on the inner surface of the apoferritin cage, resulting in variation in the shape of the core, as seen for Fe-ferritin. In contrast to these salts conditions, indium cores are formed at low pH (about 2.8), where acidic amino acids do not dissociate. The mechanism of In core formation therefore differs from that for other salts and, although the mechanism has yet to be clarified, evidently promotes the efficient uptake of In and the formation of uniform cores.

Nano Lett., Vol. 5, No. 5, 2005

Acknowledgment. The authors thank K. Iwahori, R. Tsukamoto, and I. Yamashita for helpful discussions on recombinant apoferritin. Matsushita Electric Industrial Co. is acknowledged for financial support in the initial stages of this project. This work was supported in part by the “HighTech Research Center” project for private universities; matching fund subsidy from the Ministry of Education, Culture, Sports, Science and Technology of Japan. References (1) Meldrum, F. C.; Wade, V. J.; Nimmo, D. L.; Heywood, B. R.; Mann, S. Nature 1991, 349, 684-685. (2) Hari, A.; Ceci, P.; Ferrari, D.; Luigi, G. L.; Chiancone, E. J. Biol. Chem. 2002, 277, 37619-37623. (3) Douglas, T.; Young, M. Nature 1998, 393, 152-155. (4) Yoshimura, H.; Scheybani, T.; Baumeister, W.; Nagayama, K. Langmuir 1994, 10, 3290-3295. (5) Loewus, M. W.; Fineberg, R. A. Biochim. Biophys. Acta 1957, 26, 441-443. (6) Harrison, P. M.; Fischbach, F. A.; Hoy, T. G.; Haggis, G. H. Nature 1967, 216, 1188-1190. (7) Douglas T.; Stark V. T. Inorg. Chem. 2000, 39, 1828-1830. (8) Okuda, M.; Iwahori, K.; Yamashita, I.; Yoshimura, H. Biotechnol. Bioeng. 2003, 84, 187-194. (9) Takeda, S.; Yoshimura, H.; Endo, S.; Takahashi, T.; Nagayama, K. Proteins 1995, 23, 548-556. (10) Takeda S.; Ohta M.; Ebina S.; Nagayama K. Biochim. Biophys. Acta 1993, 1174, 218-220. (11) Macara, I. G.; Hoy, T. G.; Harrison, P. M. Biochem. J. 1972, 126, 151-162. (12) Yamashita, I. Thin Solid Films 2001, 393, 12-18.

NL050556Q

993