Programmable Assembly of Nanoarchitectures Using Genetically

NANO LETTERS

Programmable Assembly of Nanoarchitectures Using Genetically Engineered Viruses

2005 Vol. 5, No. 7 1429-1434

Yu Huang,†,‡,∇ Chung-Yi Chiang,†,∇ Soo Kwan Lee,§ Yan Gao,| Evelyn L. Hu,|,⊥ James De Yoreo,# and Angela M. Belcher*,†,§ Department of Materials Science and Engineering and DiVision of Biological Engineering, Massachusetts Institute of Technology, 77 Massachusetts AVenue, Cambridge, Massachusetts 02139, Directorate of Chemistry and Materials Science and Laboratory Science and Technology Office, Lawrence LiVermore National Laboratory, LiVermore, California 94551, and Department of Materials and Department of Electrical and Computer Engineering, UniVersity of California, Santa Barbara, California 93106 Received April 28, 2005; Revised Manuscript Received May 14, 2005

ABSTRACT Biological systems possess inherent molecular recognition and self-assembly capabilities and are attractive templates for constructing complex material structures with molecular precision. Here we report the assembly of various nanoachitectures including nanoparticle arrays, heteronanoparticle architectures, and nanowires utilizing highly engineered M13 bacteriophage as templates. The genome of M13 phage can be rationally engineered to produce viral particles with distinct substrate-specific peptides expressed on the filamentous capsid and the ends, providing a generic template for programmable assembly of complex nanostructures. Phage clones with gold-binding motifs on the capsid and streptavidin-binding motifs at one end are created and used to assemble Au and CdSe nanocrytals into ordered one-dimensional arrays and more complex geometries. Initial studies show such nanoparticle arrays can further function as templates to nucleate highly conductive nanowires that are important for addressing/interconnecting individual nanostructures.

Recent efforts focused on the bottom-up assembly of functional nanosystems from nanoscale building blocks have led to substantial advances.1-4 However, the difficulties in the precise placement and interconnection of individual building blocks1-2,5-6 have impeded further progress and thereby motivated significant efforts directed toward the controlled assembly of nanoscale building blocks.7-9 Biosystems are highly organized from molecular-scale building blocks such as nucleic acids and proteins with intricate hierarchical architectures and represent unparalleled examples of bottom-up assembly.10-11 Biological molecules often possess highly specific and precise molecular recognition * Author to whom correspondence should be addressed. E-mail: [email protected]. † Department of Materials Science and Engineering, Massachusetts Institute of Technology. ‡ Directorate of Chemistry and Materials Science, Lawrence Livermore National Laboratory. § The Division of Biological Engineering, Massachusetts Institute of Technology. || Department of Materials, University of California. ⊥ Department of Electrical and Computer Engineering, University of California. # Laboratory Science and Technology Office, Lawrence Livermore National Laboratory. ∇ These authors contributed equally to this work. 10.1021/nl050795d CCC: $30.25 Published on Web 06/03/2005

© 2005 American Chemical Society

capabilities that can be programmed through genetic engineering12-14 and can exert rational control over inorganic material nucleation, phase stabilization, assembly, and pattern formation at a molecular scale.15-19 Integration of these unique capabilities of biomolecules with the design of new materials can offer many opportunities for nanofabrication including rational molecular design through genetics, spatial control on a nanometer scale, and hierarchical assembly of two-dimensional (2D) or three-dimensional (3D) complex architectures.20-24 Here we demonstrate the assembly of various nanoachitectures including nanoparticle arrays and hetero-nanoparticle architectures utilizing highly engineered M13 bacteriophage as templates. We also show that the nanoparticle arrays can further serve as templates to nucleate highly conductive nanowires, providing a new avenue for addressing/interconnecting individual nanostructures. M13 phage is a type of filamentous virus that has an inherent structural advantage for constructing rich nanostructures. It has a single-strand DNA enclosed in a cylindrical capsid composed of approximately 2700 copies of the pVIII proteins, about 5 copies each of the pIII and pVI

Figure 1. Engineering and characterization of type 8-3 phage. (A) Scheme for engineering the type 8-3 phage. The genome of the engineered bacteriophage bears insertions in gVIII and gIII, which leads to motif expressions on pVIII and pIII proteins, respectively. Gene insertions and the correspondingly expressed motifs are highlighted in yellow and red. This engineered phage can further template assembly of a variety of nanoarchitectures. (B-C) Experimental proofs of expressed peptide motifs on clone #9s1. (B) ELISA plate result of the interactions between the s1 motif and the streptavidin-coated wells. The green color indicates positive interaction. The intensity of the green color represents the amount of the bound phage in the well. Serial dilutions (50%, 2-fold) of each phage clone were carried out in a row of ELISA plate wells. (C) The mass spectra of wild-type (M13KE bateriophage from New England Biolabs) and #9s1 clones.

proteins at one end of the bacteriophage, and about 5 copies each of the pVII and pIX proteins at the other end. The functionalities of these protein groups residing at different locations on a viral particle can be rationally altered independently via genetic engineering.25 Programmable protein functionalities promise the ultimate flexibility in assembling heterofunctional nanostructures, which makes this viral system an attractive template for synthesis and assembly of various materials and structures. However, exploring highly engineered heterofunctional viral templates for materials assembly has not been reported to date. We therefore set out to achieve engineered viral templates with desired modifications on both pIII and pVIII proteins, which will be termed the type 8-3 clone. To identify suitable peptide motifs for individual protein groups, two types of phage libraries were employed, type 3 and type 8 libraries (Figure 1A). After specific binding motifs for targeted substrates were selected from separate libraries, a type 8-3 phage was produced with different binding motifs on both pIII and pVIII proteins by combining gene III (gIII) and gene VIII (gVIII) insertions in a single viral genome (Figure 1A). The type 3 library is the most commonly used screening pool, which is based on a combinatorial library of random peptides fused to pIII proteins. However, peptide motifs screened from a type 3 library cannot be directly translated to pVIII proteins because pIII and pVIII proteins have their distinct structure information. Fusing binding motifs chosen from the type 3 phage library to pVIII proteins may cause a deviation in the functionality of the displayed motifs and vice versa.26 Therefore, to identify the desirable peptide motifs for pVIII proteins, a type 8 engineered library was exploited. 1430

The type 8 library that we used is constructed by fusing eight random amino acids into the N-terminus of all the 2700 copies of the pVIII proteins27 with a random population of 107-108. This library employs a modified M13KE phage vector (New England Biolabs, we will refer to it as the wildtype in the following studies) by generating restriction sites, PstI and BamHI, in gVIII through mutagenesis for the insertion of random codons (Gnm(nnm)6nnG), where n ) G, C, A or T and m ) T or G). Unlike the previously reported phage engineered using a phagemid system, phage in the type 8 library employ genome engineering that produces 100% expression and monodispersed viral dimensions.25 Through the use of a general biopanning technique by exposing the type 8 phage library to gold thin films, antigold binding motifs on pVIII proteins were selected. The peptide sequence Val-Ser-Gly-Ser-Ser-Pro-Asp-Ser (VSGSSPDS), which is named p8#9, emerged as the dominant binding motif after the forth round of the selection.28 This goldbinding motif contains four serines, each of which has a hydroxyl group on the side chain. Hydroxyl-rich peptides have been reported to have high affinity to gold lattices.11,29 Our other unpublished data obtained from the biopanning of the type 3 library (New England Biolabs) also show that hydroxyl-rich motifs are favorable for binding gold surfaces. In the specific type 8-3 clone reported here, the p8#9 motif was chosen for pVIII protein fusion, and a 12-mer peptide sequence (SWDPYSHLLQHPQ) was chosen for pIII protein fusion. This 12-mer peptide motif termed s1 was screened previously through a type 3 library for specific binding to streptavidin.30 We hence name this new type 8-3 clone #9s1, the genome of which bears insertions on both gIII and gVIII. Nano Lett., Vol. 5, No. 7, 2005

Figure 2. Gold-binding experiments and corresponding TEM images. (A) The photograph of different clones mixed with Au nanoparticles. Only sample p8#9 shows visible precipitate, highlighted by the arrow. All the other solutions remain clear. (B) Au nanoparticles (blank control sample). (C) Wild-type phage mixed with Au nanoparticles (negative control sample), stained with 2% uranyl acetate. (D) Sample p8#17 mixed with Au nanoparticles (negative control sample), stained with 2% uranyl acetate. (E) Distinguishable wirelike structures in the mixture of sample p8#9 and Au particles. The inset shows Au particles self-assembled into 1D arrays on the virus.

We carried out experiments to demonstrate the expressions and unimpaired functionalities of both peptide motifs on pIII and pVIII proteins in this new clone. DNA sequencing experiments demonstrated the successful insertions in both gIII and gVIII of a single phage genome. The presence and binding affinity of s1 motifs on pIII proteins were verified by an enzyme-linked immunosorbent assay (ELISA)-based method. Each well of the ELISA plate is coated with streptavidin (Pierce Biotechnology) that can be bound by a s1 motif. After the wells interacted with the different phage clones for an hour, they were washed to remove unbound phage. Horseradish peroxidase (HRP)-conjugated pVIII antibodies (Amersham Biosciences) were added to bind to pVIII proteins of the viral particles that had bound to the streptavidin-coated wells. HRP turns the ABTS substrate (2, 2′-azino-bis(3-enthylbenzthiazonline-6-sulfonic acid, SigmaAldrich) from colorless to green in the presence of H2O2 (hydrogen peroxide, Sigma-Aldrich). Green coloration was only observed in the wells that were exposed to p3s1 clones (positive assay with clones demonstrated to have specific streptavidin-binding (s1) motif on pIII proteins and with wildtype pVIII proteins) and #9s1 clones, indicating that interaction had occurred between the viruses and the streptavidin Nano Lett., Vol. 5, No. 7, 2005

Figure 3. (A-E) TEM images of various nanoarchitectures templated by clone #9s1. Gold nanoparticles (∼5 nm) bind to pVIII proteins along the virus axis and form 1D arrays, while s1 motif on pIII protein simultaneously binds to streptavidin-coated nanoparticles. Arrows highlight the streptavidin-conjugated gold nanoparticles (∼15 nm) and CdSe quantum dots bound on pIII proteins. The insets show the assembly schemes of observed structures. White represents the virus structure, yellow dots represent gold nanoparticles, the green dot represents a CdSe quantum dot, and red represents the streptavidin coating around gold or CdSe particles. (C, inset) The enlarged image of the CdSe quantum dot attached to the end of the virus.

coating of the wells (Figure 1B). The adsorption data shows that p3s1 and #9s1 have similar binding affinities to streptavidin. For the wells exposed to p8#9 clones (negative assay with clones with wild-type pIII proteins and with the p8#9 gold-binding motif on pVIII proteins), no color change was observed, indicating no interaction had occurred, just as in the blank assay control (Figure 1B). The expression of the octapeptide motifs on pVIII proteins was proven by mass spectroscopy. The mass spectrum of the phage proteins was acquired using an Applied Biosystems model Voyager DESTR instrument for matrix assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry operating in linear mode. In the mass spectrum of wild-type M13 bacteriophage (M13KE), the maximum peak at 5252.06 Da corresponds to the molecular weight of native pVIII proteins in M13KE. The peak at 5655.25 Da in the mass spectrum of #9s1 corresponds to the molecular weight of the engineered pVIII proteins, indicating that the anti-gold motif is 100% displayed on the pVIII proteins of bacteriophage (Figure 1C). 1431

Figure 4. (A-C) TEM images of the progressive growth of continuous gold nanowires templated by gold nanoparticle arrays on pVIII proteins: (A) before electroless deposition, (B) 3-min deposition, and (C) 5-min deposition. (D) TEM image of gold nanowires obtained via direct nucleation of gold from solution (5 mM chloroauric acid + 5 mM sodium borohydride). The inset shows an enlarged view of a nucleated gold wire. (E) Three-dimensional plot of an atomic force microscopy (AFM) image of a two-terminal device based on a single viral gold nanowire as described in part C. (F) The two-terminal I-V behavior of a nanowire measured at room temperature.

With the successful modifications on both pIII and pVIII proteins, type 8-3 phage can serve as a template for assembly of nanoarchitectures with rich geometries and functionalities. The viral capsid with expressed anti-gold motifs on pVIII proteins can serve as a template for the organization of one-dimensional (1D) Au nanocrystal arrays. To demonstrate the specific templating effect of p8#9 motifs on pVIII, experiments were carried out with different clones, including p8#9 (positive motif) and negative controls wildtype and p8#17 (a random motif, GLDGSAVD, selected from the type 8 library). Selected clones were all amplified to the concentration of ∼108 pfu/µL. Approximately 10µL of the phage solutions were mixed with 100 µL of 5-nm gold colloidal suspension (5.0 × 1013 particles/mL, Ted Pella). After 2-3 min of incubation, the mixtures of control clones with the gold colloidal suspension and the blank control gold suspension remained clear (Figure 2A). Transmission electron microscopy (TEM) studies on these samples showed that gold particles distributed randomly (Figures 2BD). Uranyl-acetate-stained samples showed clearly that Au nanoparticles were not templated by either wild-type phage or p8#17 clones (Figures 2C-D). In the case of the p8#9 motif, a visible precipitate in the solution was observed, indicating that gold colloids formed aggregates. TEM studies showed that gold nanoparticles assembled into wirelike structures with diameters around 10 nm and lengths around 1 µm, which is consistent with the morphology of the M13 virus (Figure 2E). An enlarged image revealed that gold nanoparticles were arranged into an ordered 1D array on pVIII proteins along the viral axis (Figure 2E, inset). Onedimensional metallic nanoparticle arrays may serve as the 1432

basis for plasmon structures that can selectively absorb or transmit specific electromagnetic radiation with high spectral sensitivity.31 The #9s1 clones can further be utilized to assemble heterostructures with Au particles forming 1D arrays on pVIII proteins and another streptavidin-coated material bound to pIII proteins. The #9s1 clone was first amplified to 109 pfu/µL. Approximately 10 µL of the phage solution was mixed with 5 µL of 15-nm gold-streptavidin conjugates (Ted Pella) and 70 µL of TBST (Tris-buffered saline (TBS) + 0.05% (v/v) Tween-20 + 0.1% bovine serum albumin (BSA, Amersham Biosciences)), and the mixture was agitated for about 1 h, allowing s1 on the pIII proteins to bind to streptavidin-coated Au. Then, 10 µL of the mixture solution was incubated with 100 µL of 5-nm colloidal Au (Ted Pella) to allow gold nanoparticles to bind to pVIII proteins. The resulting structures were examined with TEM. The images revealed that 5-nm Au particles formed 1D arrays on pVIII proteins and the streptavidin-coated 15-nm gold particles bound to one end of the phage as a result of functionalities on pIII proteins (Figures 3A-B). This rational approach provides the assembly of nanostructures based on genetic engineering of the viral system and can easily be extended to other materials/structures. For example, when streptavidin-coated Au was replaced with streptavidinconjugated cadmium selenide (CdSe) quantum dots (Quantum Dot), hetero-nanoparticle arrays consisting of CdSe quantum dots (QDs) bound to pIII proteins and gold nanoparticle arrays bound to pVIII proteins were obtained (Figure 3C). More than one phage can connect to the same streptavidin-coated nanoparticle and create more complex Nano Lett., Vol. 5, No. 7, 2005

structures, such as a linear wire-dot-wire constructs or a Y-shaped structure (Figures 3D-E). Moreover, the 1D metallic nanoparticle arrays on pVIII proteins can subsequently grow into continuous metallic nanowires. This was achieved by electroless deposition of gold onto the existing Au nanoparticle arrays using the catalytic effects of gold surfaces to selectively reduce gold from HAuCl4 (0.3 mM, chloroauric acid, Sigma-Aldrich) in the presence of NH2OH (4 mM, hydroxylamine, Sigma-Aldrich).32 Growth of the 1D nanoparticle arrays into continuous nanowires was verified by TEM studies, which revealed a progressive growth of the nanowires with increasing electroless deposition time. By control of the electroless deposition time, continuous Au nanowires can be achieved (Figures 4A-C). Continuous nanowires obtained after 5 min of deposition have an average diameter of ∼40 nm. Gold nanowires can also be obtained from direct nucleation from gold salt solution onto the pVIII proteins of p8#9 viruses (Figure 4D). Control experiments using M13KE and p8#17, the same clones used for control of binding experiments in Figure 2, did not promote gold wire formation (Supporting Information). These nanoscale metal wires (templated on pVIII proteins) can be used to route electrical carriers to a nanostructure (templated on pIII proteins) in a precise and controlled way, which provides an approach to solving the interconnecting problem in the studies of quantum objects.1-2,6 To further investigate the potential of these biotemplated nanowires as electronic components in future nanoelectronic circuitry, electrical transport studies were performed on individual nanowires. Electron-beam lithography was used to make electrical contacts to the nanowires (Figure 4E). Two-terminal current-voltage (I-V) measurements were conducted at room temperature. For a continuous nanowire, the current-voltage curve shows a linear relation (Figure 4F), indicating the ohmic contact behavior. Repeated measurements on the nanowire yielded reproducible I-V behaviors, indicating that the device was stable. A resistance of about 588 Ω is estimated for the device. These data give a resistivity of ∼1.8 × 10-6 Ωm for the gold nanowire, which is about 100 times larger than that of bulk gold (2.24 × 10-8 Ωm). Diameter fluctuation, surface scattering, and tunneling through different nanocrystal grain boundaries may contribute to this result. With an increase in bias voltage, the device failed when the current reached around 500 µA, which yielded a high failure current density about 108 A/cm2. More thorough characterization of the phage-templated wires is underway and will relate the nanostructure of the wires to the corresponding electrical performance. We have demonstrated that type 8-3 engineered phage can serve as templates to assemble nanoarchitectures of various geometries and materials, such as Au and CdSe nanocrystal/hetero-nanocrystal arrays and Au nanowires. These results suggest a generic assembly approach that can be extended to program the assembly of various material systems with nanometric precision; various substrate-specific motifs can be independently selected from type 8 and type 3 libraries and then genetically incorporated into M13 structures to produce highly decorative viral templates with Nano Lett., Vol. 5, No. 7, 2005

rational control. Overall, this work demonstrates the great promise of bottom-up assembly for sophisticated electronic and optoelectronic systems from biotemplated materials and therefore opens up many exciting opportunities in nanoscale science and biotechnology. Acknowledgment. This work was supported by the Army Research Office Institute of Collaborative Biotechnologies, Microelectronics Advanced Research Corporation, and its Focus Center on Function Engineered Nanoarchitectonics. Y.H. acknowledges the Lawrence Livermore National Laboratories for fellowship support. Supporting Information Available: Gold nucleation experiments and TEM images. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Liang, W. J.; Shores, M. P.; Bockrath, M.; Long, J. R.; Park, H. Nature 2002, 417, 725-729. (2) Klein, D. L.; Roth, R.; Lim, A. K. L.; Alivisatos, A. P.; McEuen, P. L. Nature 1997, 389, 699-701. (3) Huang, Y.; Duan, X. F.; Cui, Y.; Lauhon, L. J.; Kim, K. H.; Lieber, C. M. Science 2001, 294, 1313-1317. (4) Tans, S. J.; Verschueren, A. R. M.; Dekker, C. Nature 1998, 393, 49-52. (5) Keren, K.; Berman, R. S.; Buchstab, E.; Sivan, U.; Braun, E. Science 2003, 302, 1380-1382. (6) Deshmukh, M. M.; Prieto, A. L.; Gu, Q.; Park, H. Nano Lett. 2003, 3, 1383-1385. (7) Yang, P. D.; Kim, F. ChemPhysChem 2002, 3, 503-506. (8) Huang, Y.; Duan, X. F.; Wei, Q. Q.; Lieber, C. M. Science 2001, 291, 630-633. (9) Alivisatos, A. P. Science 2000, 289, 736-737. (10) Seeman, N. C.; Belcher, A. M. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 6451-6455. (11) Sarikaya, M.; Tamerler, C.; Jen, A. K. Y.; Schulten, K.; Baneyx, F. Nat. Mater. 2003, 2, 577-585. (12) Brown, S. Proc. Natl. Acad. Sci. U.S.A. 1992, 89, 8651-8655. (13) Whaley, S. R.; English, D. S.; Hu, E. L.; Barbara, P. F.; Belcher, A. M. Nature 2000, 405, 665-668. (14) Nam, K. T.; Peelle, B. R.; Lee, S. W.; Belcher, A. M. Nano Lett. 2004, 4, 23-27. (15) Douglas, T.; Young, M. Nature 1998, 393, 152-155. (16) Mao, C.; Solis, D. J.; Reiss, B. D.; Kottmann, S. T.; Sweeney, R. Y.; Hayhurst, A.; Georgiou, G.; Iverson, B.; Belcher, A. M. Science 2004, 303, 213-217. (17) Hartgerink, J. D.; Beniash, E.; Stupp, S. I. Science 2001, 294, 16841688. (18) Reches, M.; Gazit, E. Science 2003, 300, 625-627. (19) Weizmann, Y.; Patolsky, F.; Popov, I.; Willner, I. Nano Lett. 2004, 4, 787-792. (20) Yan, H.; Park, S. H.; Finkelstein, G.; Reif, J. H.; LaBean, T. H. Science 2003, 301, 1882-1884. (21) Li, Z.; Chung, S. W.; Nam, J. M.; Ginger, D. S.; Mirkin, C. A. Angew. Chem., Int. Ed. 2003, 42, 2306-2309. (22) Dujardin, E.; Peet, C.; Stubbs, G.; Culver, J. N.; Mann, S. Nano Lett. 2003, 3, 413-417. (23) Hall, S. R.; Shenton, W.; Engelhardt, H.; Mann, S. ChemPhysChem 2001, 2, 184-186. (24) Lee, S. W.; Mao, C.; Flynn, C. E.; Belcher, A. M. Science 2002, 296, 892-895. (25) Armstrong, N.; Adey N. B.; McConnell S. J.; Kay B. K. Phage Display of Peptides and Proteins: A Laboratory Manual; Kay B. K., Winter J., McCafferty J., Eds.; Academic Press: San Diego, 1996, pp 35-53. (26) Matsuura, T.; Miyai, K.; Trakulnaleamsai, S.; Yomo, T.; Shima, Y.; Miki, S.; Yamamoto, K.; Urabe, I. Nat. Biotechnol. 1999, 17, 5861. (27) Petrenko, V. A.; Smith, G. P.; Gong, X.; Quinn, T. Protein Eng. 1996, 9, 797-801. 1433

(28) The library in Tris-buffered saline (TBS, pH 7.5) containing 0.1% (v/v) Tween-20 was exposed to the gold-coated slide (EMF). After the gold-coated slide was rocked for 1 h, it was washed 10 times with TBS (pH 7.5). The bound phage were eluted by addition of 1 mL glycine-HCl (0.2M, pH 2.2) containing 1 mg/mL bovine serum albumin (BSA). The eluate was neutralized by adding 300 µL of 1 M Tris-HCl (pH 9.1). A small amount (∼10 µL) of the eluate was plated on LB XGal/IPTG plates. Plaques on the plates were randomly picked, and the DNA was sequenced. The rest of the eluate was mixed with an Escherichia coli ER2537 (New England Biolabs) host to

1434

(29) (30) (31) (32)

amplify the library eluate to the concentration of about 109 pfu/µL. This amplified sublibrary was used for the next round biopanning. Brown, S. Nat. Biotechnol. 1997, 15, 269-272. Lee, S. W.; Lee, S. K.; Belcher, A. M. AdV. Mater. 2003, 15, 689692. Brongersma, M. L.; Hartman, J. W.; Atwater, H. A. Phys. ReV. B 2000, 62, R16356-R16359. Brown, K. R.; Natan, M. J. Langmuir 1998, 14, 726-728.

NL050795D

Nano Lett., Vol. 5, No. 7, 2005