Molecular Recognition-Mediated Fabrication of Protein

Biologically Inspired Materials and Materials Systems, Duke UniVersity. Received August 5 ... terminal COOH groups in the MHA SAM were activated by re...
0 downloads 0 Views 373KB Size
Download PDF
NANO LETTERS

Molecular Recognition-Mediated Fabrication of Protein Nanostructures by Dip-Pen Lithography

2002 Vol. 2, No. 11 1203-1207

Jinho Hyun,† Sang Jung Ahn,‡ Woo Kyung Lee,‡ Ashutosh Chilkoti,*,†,§ and Stefan Zauscher*,‡,§ Department of Biomedical Engineering, Duke UniVersity, Box 90281, Durham, North Carolina 27708-0281, Department of Mechanical and Materials Science, Duke UniVersity, Box 90300, Durham, North Carolina 27708-0300, and Center for Biologically Inspired Materials and Materials Systems, Duke UniVersity Received August 5, 2002; Revised Manuscript Received September 10, 2002

ABSTRACT We describe the molecular recognition-mediated, stepwise fabrication of patterned protein nanostructures with feature sizes on the order of 200 nm. First, a self-assembled monolayer (SAM) of 16-mercaptohexadecanoic acid (MHA) is patterned onto gold by dip-pen nanolithography (DPN), and the unpatterned regions are passivated with a protein-resistant oligoethylene glycol-terminated alkanethiol SAM. Next, an amineterminated biotin derivative is covalently conjugated with the chemically activated MHA SAM nanopattern. The surface is then incubated with streptavidin to form streptavidin nanostructures, mediated by molecular recognition between biotin and streptavidin. Finally, protein nanopatterns are fabricated by molecular recognition-mediated immobilization of biotinylated protein from solution. Our fabrication methodology is generically applicable because of the ubiquity of biotin-tagged molecules.

Introduction. The spatially controlled immobilization of biomolecules on solid surfaces on the nanometer length scale is driven by the possibility of fabricating protein nanoarrays with well-defined feature size, shape, and spacing. Such structures are important for the fundamental study of the interactions between cells and surfaces1,2 and have potential applications in the design of cell-based sensors and biomaterials.3,4 This paper describes how dip-pen nanolithography (DPN)5-7 in combination with the high-affinity streptavidinbiotin protein-ligand system8,9 provides a simple and versatile “bottom-up” approach to create nanoscale biomolecular structures readily in a stepwise fashion. Our method involves the fabrication of chemically reactive nanoscale features by patterning a self-assembled monolayer (SAM) of a COOH-terminated alkanethiol on a gold substrate by DPN, followed by covalent immobilization of a high-affinity small-molecule ligand (biotin) on the nanopatterned SAM and subsequent molecular recognition of its protein-binding partner (streptavidin) from solution. The resulting streptavidin nanopattern provides a universal platform for molecular recognition-mediated protein immobilization because of * Corresponding authors. E-mail: [email protected]. [email protected]. † Department of Biomedical Engineering, Duke University. ‡ Department of Mechanical and Materials Science, Duke University. § Center for Biologically Inspired Materials and Materials Systems. 10.1021/nl0257364 CCC: $22.00 Published on Web 09/28/2002

© 2002 American Chemical Society

the ubiquity of biotin-tagged molecules. Typical periodic arrays of biotin-BSA fabricated by our method are shown in Figure 1. Experimental Section. Gold substrates with an average Au grain diameter of 30 nm were prepared by thermal evaporation of a chromium adhesion layer (100 Å), followed by gold deposition (1000 Å) onto a glass cover slide at 4 × 10-7 Torr. Before deposition, the glass surface was cleaned in a 5:1:1 (v/v/v) mixture of H2O, H2O2, and NH3 at 80 °C for 20 min. MHA (16-mercaptohexadecanoic acid) was patterned on the gold surface with DPN using an atomic force microscope (AFM) (MultiModeTM, Digital Instruments). For this purpose, an AFM cantilever (silicon nitride cantilever, 0.05 N/m, Digital Instruments) was incubated in a saturated solution of MHA in degassed acetonitrile for 1 min. The relative humidity during patterning ranged from 35 to 55%. Patterns were generated with writing speeds of up to 8 µm/s and resulted in feature widths of less than 100 nm that could be resolved with lateral force microscopy. MHA SAM arrays with periodic features (feature sizes ranging from 100 to 2000 nm) were routinely patterned by programming the XY motion of the AFM tube scanner through a customized nanolithography program (NanoScriptTM, Digital Instruments). Patterned areas could be located accurately and repeatedly by pixel

Scheme 1 Stepwise Fabrication Process to Create Molecular Recognition-Mediated Protein Nanostructures

Figure 1. Biotin-BSA nanopatterns imaged with AFM TappingModeTM: (A) (i) 144-dot pattern with an average feature size of 230 nm; (ii) zoom of the area indicated by the yellow square in A(i) and the representative cross section showing the typical feature size; (B) 9-dot pattern with about a 1-µm feature size.

correlation using still-video micrographs captured during lithography. To prevent nonspecific protein adsorption in subsequent processing steps, the bare gold substrate background was passivated by incubation with an ethanolic solution of 11mercaptoundecyl-tri(ethylene glycol) (EG3-SH) for 1 h to form a protein-resistant, “nonfouling” SAM surface.4,10 The terminal COOH groups in the MHA SAM were activated by reaction with N-hydroxysuccinimide (NHS) (0.2 M Aldrich) and 1-ethyl-3-(dimethylamino)propyl carbodiimide (EDAC) (0.1 M Aldrich) in deionized water for 30 min. Before further processing, the samples were sonicated in ethanol for 5 min, rinsed with ethanol and deionized water, and dried in a stream of N2 gas. To couple biotin to the surface covalently, a 200-µL droplet of a 10 mM solution of biotinyl-3,6,9-trioxaundecanediamine (biotin-NH2, Pierce) in phosphate buffer (pH 7.4) was pipetted onto the patterned area and left for 10 min. Fluorescently labeled streptavidin (Alexa488 conjugate, Molecular Probes) was bound to the biotin pattern by molecular recognition during a 10-min incubation period at room temperature (0.1 µM streptavidin in 10 mM HEPES containing 0.02% (v/v) Tween 20 detergent). After surface modification, samples were exhaustively washed in PBS buffer and deionized water and were immediately imaged by AFM. Biotinylated BSA (biotin-BSA, Molecular Probes) was prepared by dissolving 2 mg BSA in 1 mL of 50 mM sodium bicarbonate buffer at pH 8.0 and subsequently adding 75 1204

µL of NHS-LC-biotin (EZ-Link NHS-LC-biotin, Pierce) (2 mM, DMF) to the BSA solution and incubating for 2 h. Unincorporated biotin was removed by gel filtration (PD10 column, Pharmacia Biotech). BSA nanopatterns were generated by incubating the streptavidin pattern for 10 min with 0.1 µM biotin-BSA in 10 mM HEPES containing 0.02% (v/v) Tween 20 detergent at room temperature. The feature height after each fabrication step was determined from averaged cross-sectional profiles of AFM height images, and the patterns of fluorescently labeled streptavidin were observed with confocal fluorescence microscopy (LSM 510, Zeiss). Results and Discussion. Here we describe the stepwise fabrication of protein nanopatterns. First, a gold substrate was patterned with a SAM of MHA using DPN (Scheme 1, Step 1; Figure 2A). Patterns could be resolved by lateral force microscopy, and line widths of less than 100 nm could be achieved routinely. We found that the minimum achievable feature size depended significantly on the roughness of the gold substrate. The bare gold background of the surface was passivated by incubation with EG3-SH (Scheme 1, Step 2) to create a protein-resistant nonfouling SAM in the unpatterned region. This critical processing step prevented nonspecific protein adsorption in subsequent processing steps.4,10 After passivation of the background with the EG3-SH SAM, the AFM height image of the patterned regions showed negative contrast against the background, resulting from the longer chain length of EG3-SH as compared to that of MHA (Figure 2E). Next, the terminal COOH groups in the patterned MHA SAM were converted to a reactive Nhydroxysuccinimide ester, followed by reaction with biotinNH2, thus covalently binding biotin to the patterned regions (Scheme 1, Step 3). An AFM image of the surface after this step (Figure 2B) revealed an ∼0.8 nm height difference between the patterned region and the background (Figure 2E). After covalently binding biotin-NH2 to the activated MHA pattern, the substrate was incubated in a solution of Nano Lett., Vol. 2, No. 11, 2002

Figure 3. (A) (i) Height image (TappingModeTM) and average cross section of a patterned Alexa488-streptavidin dot. (ii) Confocal fluorescence image of A(i). (B) (i) Height image and average cross section of a patterned biotin dot after incubation with Alexa488streptavidin preincubated with excess biotin. (ii) Confocal fluorescence image of B(i). Figure 2. (A) Lateral force image of an MHA SAM on gold patterned by DPN. (B) Height image (TappingModeTM) of biotinamine coupled to the MHA SAM after backfilling the background in (A) with an EG3-SH SAM and activation of the terminal COOH groups of the MHA SAM. (C) Height image (TappingModeTM) of the same surface after incubation of the biotin pattern in (B) with streptavidin. (D) Height image (TappingModeTM) of the surface after incubation of the streptavidin pattern in (C) with biotinylated BSA. (E) Height difference between the pattern and the background referenced to the background EG3-SH SAM in (B) (labels A-D refer to images A-D).

fluorescently labeled streptavidin (Scheme 1, Step 4; Figure 2C). A significant increase in the feature height after incubation with streptavidin (Figure 2E) suggested that streptavidin was preferentially immobilized on patterned regions of the surface. Larger, micrometer-size circular patterns of fluorescently labeled streptavidin were fabricated and imaged by AFM (Figure 3A(i)) and by confocal fluorescence microscopy (Figure 3A(ii)) to corroborate this observation independently. Fluorescence microscopy confirmed that the streptavidin bound selectively to the biotin pattern on the surface (Figure 3A(ii)). To verify that the formation of streptavidin patterns was caused by molecular recognition of the surface-bound biotin through solution-phase streptavidin and not by nonspecific adsorption of the protein, a control experiment wherein a biotin pattern was incubated with 0.1 mM Alexa488-streptavidin that had been previously incubated with 100 mM biotin to block all available biotin binding Nano Lett., Vol. 2, No. 11, 2002

sites was performed. The average feature height obtained from the height profile of the AFM image (Figure 3B(i)) was similar to that of the biotin-patterned surface before incubation, and fluorescence microscopy showed the absence of streptavidin (Figure 3B(ii)). Both observations suggest that the biotin-bound streptavidin did not bind to the patterned biotin on the surface and confirms the specificity of the streptavidin immobilization in the patterning process. The streptavidin receptor density can potentially be varied by patterning mixed monolayers of MHA and EG3-SH. We chose to fabricate biotin-streptavidin patterns because they provide a universal platform for patterning other biomolecules. This is because streptavidin, by virtue of its homotetrameric structure and dyad-related symmetry, functions as a biomolecular adapter.11 The attachment of streptavidin to surface-immobilized biotin leaves two biotin binding sites on the solution-exposed face of the protein unoccupied, which can be used to subsequently pattern other molecules that are conjugated to biotin. This modular patterning scheme is attractive because it affords a high degree of flexibility stemming from an ever-increasing supply of biotin-linked reagents. To demonstrate the ability to build upon the streptavidin nanostructures and thereby immobilize other biomolecules of interest, a streptavidin pattern was incubated with bovine serum albumin (BSA) that was conjugated with biotin (biotin-BSA) (Scheme 1, Step 5). After incubation, AFM images showed enhanced contrast (Figure 2D) when com1205

Figure 4. (A) Average feature height of a streptavidin pattern before and after incubation with BSA. (B) Average feature height of biotin-blocked and unblocked streptavidin patterns before and after incubation with biotin-BSA. Legend: SA ) streptavidin pattern. SA + BSA ) streptavidin pattern incubated with BSA. Blocked SA + Biotin-BSA ) streptavidin pattern previously blocked with excess biotin from solution and then incubated with biotin-BSA. SA + Biotin-BSA ) unblocked streptavidin pattern incubated with biotin-BSA. Feature heights were obtained from averaged cross sections of AFM TappingModeTM height images.

pared to the streptavidin pattern (Figure 2C), caused by the height increase of the patterned feature (Figure 2E). Incubation of a streptavidin pattern with BSA (not biotinylated) did not cause a significant height increase (Figure 4A). These results suggested that biotin-BSA bound specifically to the streptavidin pattern, likely by molecular recognition of streptavidin through the biotin moiety, and that unspecific adsorption of BSA was not responsible for the observed height increase. To confirm this molecular recognitionmediated, specific binding of biotin-BSA to the streptavidin pattern, we incubated biotin-BSA with a streptavidin pattern previously blocked with excess biotin and imaged the resulting pattern with AFM. The analysis of height crosssections of AFM images before and after incubation showed no significant increase in height for the blocked streptavidin pattern, whereas the unblocked streptavidin pattern showed a significant increase in height after the incubation with biotin-BSA (Figure 4B). These experiments with biotin-BSA 1206

demonstrate, as before with BSA (Figure 4A), that unspecific protein adsorption is not responsible for the observed increase in height and confirm the specificity of the binding between the biotin-BSA and immobilized streptavidin. Periodic arrays of biotin-BSA with a feature size of ∼230 nm were readily fabricated by our method, as shown in Figure 1A. The feature size and feature spacing of the patterns could be controlled by adjusting the patterning parameters such as tip-surface contact time, concentration of the MHA SAM, and humidity. Nonspecific adsorption of proteins on the background, a critical problem in the fabrication of protein arrays, was overcome by passivating the background with an EG3-SH SAM. In summary, the use of the streptavidin-biotin system as shown here, or other biomolecular recognition systems, provides a complementary approach to the fabrication of biomolecular structures by DPN that uses physisorption or covalent conjugation4,12 with some important distinctions: because the binding of the final target molecule is mediated by a highly specific molecular recognition interaction that occurs solely in the patterned region against a nonfouling background, this approach should be amenable to patterning a biomolecule of interest from a complex mixture. In particular, our nanofabrication approach may be especially useful for immobilizing recombinant proteins and peptides because they can be appended with affinity tags such as oligohistidine,13 biotin,14 and S-peptide15,16 during expression or by post-translational modification in vivo. This also raises the intriguing possibility that nanostructures of recombinant proteins and peptides could be created directly on a surface by affinity capture from cell lysate without the necessity of intermediate purification steps, which are required for alternative DPN methods that involve physisorption or covalent immobilization of the protein or peptide. Furthermore, genetically encoded affinity tags can be placed on a protein scaffold with molecular precision, enabling control of the orientation of the immobilized protein. Acknowledgment. We thank Professor Buddy Ratner (University of Washington, Seattle) for the generous gift of the EG3-thiol. A.C. acknowledges the financial support of the National Science Foundation through grant BES-9986477-NANOSCALE. S.Z. acknowledges the financial support of the American Chemical Society through grant PRF37012-G9 and the Ralph E. Powe Jr. Faculty Enhancement Award from Oak Ridge Associated Universities. References (1) Chen, C. S.; Mrksich, M.; Huang, S.; Whitesides, G. M.; Ingber, D. G. Science (Washington, D.C.) 1997, 276, 1425-1428. (2) Maheshwari; G.; Brown, G.; Lauffenburger, D. A.; Wells, A.; Griffith, L. G. J. Cell Sci. 2000, 113, 1677-1686. (3) Blawas, A. S.; Reichert, W. M. Biomaterials 1998, 19, 595-609. (4) Lee, K.-B.; Park, S.-J.; Mirkin, C. A.; Smith, J. C.; Mrksich, M. Science (Washington, D.C.) 2002, 295, 1702-1705. (5) Piner, R.; Zhu, J.; Xu, F.; Hong, S.; Mirkin, C. A. Science (Washington, D.C.) 1999, 283, 661-663. (6) Hong, S.; Zhu, J.; Mirkin, C. A. Science (Washington, D.C.) 1999, 286, 523-525. (7) Hang, S. H.; Mirkin, C. A. Science (Washington, D.C.) 2000, 288, 1808-1811. Nano Lett., Vol. 2, No. 11, 2002

(8) Chilkoti, A.; Tan, P.; Stayton, P. S. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 1754-1758. (9) Hyun, J.; Chilkoti, A. J. Am. Chem. Soc. 2001, 123, 6943-6944. (10) Prime, K. L.; Whitesides, G. M. Science (Washington, D.C.) 1991, 252, 1164-1167. (11) Weber, P. C.; Olendorf, D. H.; Wendoloski, J. J.; Salemme, F. R. Science (Washington, D.C.) 1989, 243, 85-88. (12) Liu, G.-Y.; Amro, N. A. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 5165-5170.

Nano Lett., Vol. 2, No. 11, 2002

(13) Smith, M. C.; Furman, T. C.; Ingolia, T. D.; Pidgeon, C. J. Biol. Chem. 1988, 263, 7211-7215. (14) Tsao, K. W.; deBarbieri, B.; Hanspeter, M.; Waugh, D. W. Gene 1996, 169, 59-64. (15) Smith, P. A.; Tripp, B. C.; DiBlasio-Smith, E. A.; Lu, Z.; LaValle, E. R.; McCoy, J. A. Nucleic Acids Res. 1998, 26, 1414-1420. (16) Kim, J.-S.; Raines, R. T. Protein Sci. 1993, 2, 348-356.

NL0257364

1207