NANO LETTERS
Enzyme-Assisted Nanolithography Leif Riemenschneider, Sven Blank, and Manfred Radmacher*
2005 Vol. 5, No. 9 1643-1646
Institute of Biophysics, UniVersity of Bremen, 28359 Bremen, Germany Received September 21, 2004; Revised Manuscript Received July 20, 2005
ABSTRACT We have chemically immobilized alkaline phosphatase molecules onto the apex of a tip of an atomic force microscope. When the substrate BCIP is dephosphorylated by alkaline phosphatase, it will precipitate in the presence of NBT. By bringing the tip in the vicinity of a suitable sample, we could locally deposit this complex on the sample. Thus we combined the activity of an enzyme with the accuracy in positioning a tip in scanning probe microscopy to demonstrate a novel technique referred to as enzyme-assisted nanolithography. By use of other enzymes, this method will open the possibility to chemically modify surfaces on a nanometer scale.
Future applications in nano-biotechnology will ask for techniques that allow surface structure modification on a nanometer scale. Developing a versatile and flexible technique for this type of modifications will be essential for potential functional materials. This includes the design of nanoreactors as the ultimate stage of miniaturized labs-ona-chip, of miniaturized assays,1 or of intelligent materials which could be used as smart drug-delivery systems.2 The key issue for all these applications will be a moleculardefined surface which exhibits addressability and molecular function. These requirements raise the need to locally modify molecular groups on a surface. The most versatile molecular tool kit for directed and controlled chemical modifications can be found in nature: enzymes. Thus we implemented and tested a technology which allows combining the chemical versatility of enzymes found in nature and the nanometer precision in positioning objects in scanning probe techniques such as AFM (atomic force microscopy). The technique introduced here may become an important contribution in the field of nano-biotechnology but also be the basis for alternative processes in semiconductor fabrication. In both fields, there is a constant need for decreasing structure sizes beyond the current state of the art eventually modifying surfaces at a nanometer, i.e., molecular, scale. At this length scale, structuring is equivalent with adding or removing single molecules or modifying single chemical groups. So, a bottom-up approach, where surfaces are modified in a chemical sense, e.g. by emplyoing enzymes, may be most appropriate. One possible route to implement a method for structuring surfaces at a nanometer scale is employing scanning probe techniques. For example, soft polymeric films can be modified by mechanical interaction of an atomic force microscope tip.3,4 Also the dip pen technology,5 where a * Correspondence and requests for materials should be addressed to M.R. (
[email protected]). 10.1021/nl0484550 CCC: $30.25 Published on Web 07/28/2005
© 2005 American Chemical Society
suitable ink is adsorbed on an AFM tip and then locally released on a sample, has attracted much interest. Enzymes have been used to chemically modify surfaces locally, e.g., by applying them with a micropipet to a sample coated with a substrate of this particular enzyme.6 By hydrolyzing the substrate, it became soluble and thus the surface has been modified on a length scale of several tens of micrometers. Recently enzyme molecules immobilized to an AFM tip were used to modify a sample with adsorbed enzyme molecules. Here the specific binding of the substrate molecule to the enzyme molecule has been used to physically remove the substrate from the sample.7 However, in this example the enzymatic activity itself has not been used to modify a surface. In another application phospholipase present in the bulk medium was used to modify a lipid film. Here local disturbances of the packing density of the lipid film caused by the mechanical interaction with an AFM tip were used to modify the sample locally.8 However, in all these cases surfaces coated with the enzyme’s substrate were needed. It is not clear how these schemes can be generalized to modify arbitrary surfaces. As has been introduced with the enzyme lysozyme9 and demonstrated for many other systems, an AFM tip positioned on top of an enzyme molecule can pick up conformational changes and thus monitor the activity of enzymes.10 The activity of alkaline phosphatase molecules adsorbed on a mica support (Figure 1a) can also be observed directly by atomic force microscopy (AFM). Using the substrate BCIP, which after dephosphorylation forms a precipitate in the presence of the cofactor NBT, we observed directly the activity of individual enzyme molecules adsorbed on the mica support. Little piles of precipitate were formed near active enzymes. We have modified the experimental setup such that we immobilize the enzyme molecule on the apex of an AFM tip allowing us to create deposits at arbitrary locations. So,
Figure 1. Alkaline phosphatase molecules have been adsorbed onto mica and imaged in buffer solution by AFM (a). In this image, the highest features (z-range 5 nm) correspond to intact enzyme molecules, the smaller features correspond to fragments which are always present in protein samples. When the substrate NBT is added to the buffer solution, active enzyme molecules will dephosphorylate BCIP. In the presence of NBT a precipitate is formed in the vicinity of the enzyme molecules (b). Growth of these precipitates will stop when all substrate is hydrolyzed but will start over when new substrate is added to the buffer solution (c). (In (b) and (c) the z-range has been changed to 20 nm due to larger height of aggregates compared to enzymes.)
we have combined the chemical activity of enzyme molecules and the high accuracy in positioning a tip by scanning probe microscopy to demonstrate the first implementation of enzyme-assisted nanolithography. The AFM tip was first silanized and then biotinylated using standard procedures. Alkaline phosphatase conjugated with streptavidin could be immobilized onto the tip via specific biotin-streptavidin interaction. First experiments where tip and cantilever were coated with phosphatase by immersing the tip in a phosphatase solution were not satisfactory. In this case many phosphatase molecules were active including those on the cantilever legs and even on the supporting chip. This resulted in an excessive production of precipitate which then deposited everywhere on the entire sample. Therefore we needed to design an experimental procedure which allows us to coat only the very apex of an AFM tip. We found a simple and reliable way for achieving this goal. A silicon wafer was biotinylated using the same procedure as used for coating AFM tips. Then the biotinylated wafer was incubated in a solution containing a phosphatase-streptavidin conjugate at low concentrations. This resulted in a sample sparsely populated with phosphatase molecules (about 10 per µm2) as proven by AFM imaging (Figure 2a). This sample was mounted in an AFM and imaged for about 1000 s with a biotinylated tip covering an area of 4 µm2. On average we found 40 alkaline phosphatase molecules in this field of view. Streptavidin exhibits four binding sites for biotin. For geometric reasons it is conceivable that it will bind with up to two sites to the support, exposing the other two binding sites to the AFM tip. Thus, streptavidin molecules on the sample could also bind to biotin molecules on the tip. While scanning across the surface streptavidin will bind to the tip, and one bond will be separated, either the bond between the tip and streptavidin or the bond between streptavidin and support. On average 50% of the streptavidin molecules will bind to the tip (Figure 2b-d). Since the spacer molecules used here are rather short, streptavidin molecules will only be able to bind to the very apex of the AFM tip, presumably only to the very last 5-10 nm. Thus we were able to coat an AFM tip specifically at its very apex with streptavidin phosphatase molecules. Since our experimental procedure involves the formation and forced 1644
Figure 2. AFM image (tapping mode) of single alkaline phosphatase molecules bound to a silicon wafer (a). The wafer is biotinylated and phosphatase is specifically bound to the wafer via streptavidin-biotin interactions. When the sample is imaged slowly in contact mode with a biotinylated tip (b), streptavidin will also bind to the tip (c), and consequently the very apex of the AFM tip is coated with phosphatase molecules (d).
rupture of biotin streptavidin bonds, there is a high probability for denaturing streptavidin. However, since phophatase molecules are not mechanically loaded by this process, the activity of the enzyme is maintained, as can be seen in AFM. The activity of the immobilized enzyme phosphatase was tested in a control measurement employing a standard photometric assay. We coated a silicon oxide wafer with the same procedure as used for coating tips and silicon wafers. When hydrolyzing the substrate p-nitrophenyl-phosphate (pNPP) absorption of yellow light will increase proportional to the concentration of the product p-nitrophenolate. Assuming a complete coverage of the silicon nitride wafer, we observed a maximum activity of 3200 reactions per second. The manufacturer claims an activity of 4200 reactions per second in solution, indicating that the activity of immoblized enzyme molecules is only marginally reduced. The functionalized silicon nitride wafers exhibited significant enzyNano Lett., Vol. 5, No. 9, 2005
Figure 3. The enzyme alkaline phosphatase was immobilized onto an AFM tip. The substrate BCIP was present in the surrounding medium. The product of the enzymatic reaction will form together with NBT a water-insoluble complex which will precipitate on the sample (a). Surface modification was done by two procedures: the tip was resting for 20 s in one spot, then moved rapidly to the next spot to deposit the next dot (b). Alternatively the tip was slowly moved with a velocity of 10 nm/s across the sample to form a continuous line (c). Imaging was done after the enzymatic reaction with the same tip in tapping mode.
matic activity over a course of several days. Inferring from this observation a sufficient stability of the enzyme immobilized to AFM tips can be assumed. This test of the activity of immobilized phosphatese on silicon nitride employing silane chemistry serves also as a control verifying the activity of phosphatase on the silicon nitride AFM tips. Since silicon nitride exhibits a native oxide layer, it was to be expected that binding via silane will also be possible on this support. The sample was now replaced by a piece of freshly cleaved mica to be modified and the buffer solution was replaced by a solution with the substrate BCIP and the cofactor NBT. Prior to the writing process, the surface is imaged in tapping mode to verify that it is clean. To start the actual deposition of the precipitate, the AFM tip is brought into contact with the surface by engaging it in contact mode (Figure 3a). We could form single spots by keeping the tip for a short time (20 s) in one location, then moving it to another location to form a new spot (Figure 3b). Alternatively the tip could be moved slowly across the sample (velocity 10 nm/s). This resulted in a line of deposit, which can have any arbitrary shape by steering the tip accordingly (Figure 3c). The deposit on the surface had a typical lateral dimension of 150-170 nm and a height of 10 nm. The deposits were then imaged with the same tip in tapping mode in liquids. Summary and Outlook. We have demonstrated that enzyme molecules linked to an AFM tip can be used to locally modify a sample. With the enzyme alkaline phosphatase we were able to create features of about 150 nm in Nano Lett., Vol. 5, No. 9, 2005
diameter. However, by minimizing the contact time, concentration of substrate, or number of enzyme molecules immobilized to the tip, it is conceivable to create smaller features, ultimately of molecular size. This opens the possibility for an enzyme-assisted nanolithography. Here we have demonstrated the topographic modification of the sample. It is conceivable, by employing different enzymes, to also achieve a chemical modification of the sample surface. Due to the modular setup of our enzyme immobilization scheme, it is possible to move to other enzyme systems, by using the corresponding conjugate of an enzyme with streptavidin. It is conceivable that the method described here can be used to modify samples possibly on a scale of single molecular reactions to form the basis for miniaturized devices and chemical nanoenvironments. This technique may become very important in the emerging field of nanobiotechnology. Methods. Precipitation near Adsorbed Alkaline Phosphatase. Alkaline phosphatase in buffer solution (Tris 40 mmol plus 1 mM MgCl2, pH 9.8) at a concentration of 0.02 mg/mL was incubated for 10 min on a piece of freshly cleaved mica. The sample was thoroughly flushed with pure buffer and mounted in an AFM (Nanoscope III, Digital Instruments). Imaging was performed in tapping mode in liquids using soft silicon nitride cantilevers (Oriented Twin Tips, Digital instruments, force constant 60 mN/m) at a frequency of 29 kHz. The buffer has been replaced by a buffer with BCIP and NBT at a concentration of 0.5 mmol. Functionalization of AFM Tip. Soft silicon nitride AFM cantilevers (force constant 60 mN/m) (NP-STT, Veeco Instruments, Santa Barbara, CA) were cleaned by UV irradiation and functionalized with biotin following a procedure adopted from Baselt et al. (Proc. IEEE 1997, 85 (No. 4)). The cantilevers are immersed for 30 min in a solution containing 9 mL of methanol, 370 µL of deionized water, 80 µL of concentrated acidic acid, and 230 µL of N-(2aminoethyl)(3-aminopropyl)trimethoxysilane (Sigma/Merck catalog # 8.19172.0100). Then the cantilevers are rinsed well in methanol and dried in a stream of liquid nitrogen. In the next step they are cured in an oven for 3 min at 120 °C and then immersed for 2 h in a solution of 4 mL of DMSO (Fluka/41640) containing 1 µg of NHS-Biotin (Biotin-Nhydroxysuccinimide, Sigma/H-1759). Then the cantilevers are rinsed well in ethanol and dried in a stream of nitrogen. Preparation of Alkaline Phosphatase Populated Surface. Oxidized silicon wafers (Crystec, Berlin/S 3012) were cut in square pieces (12 mm × 12 mm), cleaned for 10 min in a mixture of concentrated sulfuric acid and hydrogen peroxide (ratio 3:1) for 10 min, and then rinsed well in deionized water. They were biotinylated following the same procedure as used for modifying AFM tips described above. To populate the wafer sparsely with enzymes, a 200 µL droplet of a 0.2 nM streptavidin-phosphatase conjugate in a buffer solution is applied to the surface for 10 min. The buffer contains 40 mM TRIS and 1 mM magnesium chloride at a pH of 9.8. It proved necessary to remove enzymes that were not thoroughly immobilized by immersing the wafer 1645
in a 1 mM solution of p-nitrophenyl phosphate (PNPP, Sigma N-4665) in buffer and keeping it on a stirrer for 10 min. Verification of Immobilization of Alkaline Phosphatase. Silicon oxide wafers were cleaned and biotinylated as described above. Biotinylated wafer were incubated in a solution of 100 mmol streptavidin-phosphatase conjugate for 10 min. The sample was flushed with buffer, and then the absorbency at 405 nm was recorded using a photometer (Shimadzu UV2102-PC, Shimadzu Scientific Instruments Columbia, MD) for 25 min in the presence of the substrate pNNP at a concentration of 2.7 mM. As a control, the absorbency of a solution of alkaline phosphatese conjugated with streptavidin in the presence of pNPP was recorded. Enzyme-Assisted Nanolithography. Mica was purchased from Plano Wetzlar, Germany. The substrate solution is a 1:1 mixture of fractions A and B. Fraction A contains 40 mM TRIS and 1 mM MgCl2 at pH 9.8. The B fraction is the stock solution of the substrate BCIP/NBT as provided by the supplier (Sigma B-6404). It contains the BCIP at a concentration of 0.56 mM and NBT at a concentration of 0.48 mM. The concentration of substrate was not sufficient to produce distinct features and had to be increased by thermal evaporation of water by a factor of 5-10. Imaging and deposition were done with a commercial AFM (MFP-3D, Asylum research, Santa Barbara, CA). Images of deposits were taken in tapping mode in liquids at a drive frequency of around 30 kHz and a scan rate of 1 line per second.
1646
Acknowledgment. This work was supported by the Volkswagen Stiftung under code I/78790. We thank Monika Fritz for helpful discussions. References (1) Lynch, M.; et al. Functional protein nanoarrays for biomarker profiling. Proteomics 2004, 4, 1695-1702. (2) Chovan, T.; Guttmann, A. Microfabricated devices in biotechnology and biochemical processing. Trends Biotechnol. 2002, 20, 116-121. (3) Wendel, M.; Lorenz, H.; Kotthaus, J. P. Sharpened electron beam deposited tips for high-resolution atomic force microscope lithography and imaging. Appl. Phys. Lett. 1995, 67, 3732-3734. (4) Wendel, M.; et al. Nanolithography with an atomic force microscope. Superlattices Microstruct. 1996, 21, 1-8. (5) Piner, R. D.; Zhu, J.; Xu, F.; Hong, S.; Mirkin, C. A. “Dip-Pen” Nanolithography. Science 1999, 283, 661-663. (6) Ionescu, R. E.; Marks, R. S.; Gheber, L. A. Nanolithography using protease etching of protein surfaces. Nano Lett. 2003, 3, 1639-1642. (7) Takeda, S. et al. Lithographing of biomolecules on a substrate using an enzyme-immobilized AFM tip. Nano Lett. 2003, 3, 1471-1474. (8) Grandbois, M.; Clausen-Schaumann, H.; Gaub, H. E. Atomic force microscope imaging of phospholipid bilayer degradation by phospholipase A2. Biophys. J. 1998, 74, 2398-2404. (9) Radmacher, M.; Fritz, M.; Hansma, H. G.; Hansma, P. K. Direct observation of enzyme activity with the atomic force microscope. Science 1994, 265, 1577-1579. (10) Thomson, N. H.; Fritz, M.; Radmacher, M.; Hansma, P. K. Protein tracking and observation of protein motion using atomic force microscopy. Biophys. J. 1996, 70, 2421-2431.
NL0484550
Nano Lett., Vol. 5, No. 9, 2005
