11712
Langmuir 2007, 23, 11712-11717
Surface-Initiated Enzymatic Polymerization of DNA Dominic C. Chow†,‡ and Ashutosh Chilkoti*,†,‡ Department of Biomedical Engineering, Box 90281, and Center for Biologically Inspired Materials and Materials Systems, Box 90303, Duke UniVersity, Durham, North Carolina 27708 ReceiVed June 3, 2007. In Final Form: July 22, 2007 We describe a technique to synthesize DNA homopolymers on a surface using surface-initiated enzymatic polymerization (SIEP) with terminal deoxynucleotidyl transferase (TdTase), an enzyme that repetitively adds mononucleotides to the 3′-end of oligonucleotides. The thickness of the synthesized DNA layer was found to depend on the deoxymononucleotide monomer, in the order of dATP > dTTP . dGTP ≈ dCTP. In addition, the composition and the surface density of oligonucleotide initiators were also important in controlling the extent of DNA polymerization. The extension of single-stranded DNA chains by SIEP was further verified by their binding to antibodies specific to oligonucleotides. TdTase-mediated SIEP can also be used to grow spatially defined three-dimensional DNA structures by soft lithography and is a new tool for bioinspired fabrication at the micro- and nanoscale.
Introduction We report herein a method to synthesize DNA homopolymers on a surface by surface-initiated enzymatic polymerization (SIEP) using terminal deoxynucleotidyl transferase (TdTase). Surfaceinitiated polymerization (SIP) is of great interest for surface modifications as an alternative to conventional approaches in which a polymer is grafted from solution to a surface. The conventional “grafting-to” approach usually results in a low density of polymer chains at the surface, due to steric and thermodynamic constraints.1,2 In contrast, in SIP, a polymer is directly grown on a surface from initiator moieties that are covalently attached or chemisorbed to the surface. The primary advantage of SIP over the grafting-to approach is that it provides exquisite control over both the thickness and density of the polymer coating.3 A unique feature of SIP is that it enables the in situ synthesis of dense polymer brushes at the surface, which is not readily achieved by the “grafting-to” approach. SIP has many other useful ancillary features. First, SIP is compatible with many polymerization methods: to date, SIP has been demonstrated with free radical polymerization,3,4 atomtransfer radical polymerization (ATRP),3,5-10 reversible addition fragmentation chain transfer (RAFT) polymerization,11 living anionic and cationic polymerization,12-14 ring-opening polymeri* To whom correspondence should be addressed: Tel: 919-660-5373. Fax: 919-660-5409. E-mail:
[email protected]. † Department of Biomedical Engineering. ‡ Center for Biologically Inspired Materials and Materials Systems. (1) Sofia, S. J.; Premnath, V.; Merrill, E. W. Macromolecules 1998, 31, 5059. (2) Ikada, Y. Biomaterials 1994, 15, 725. (3) Hyun, J.; Chilkoti, A. Macromolecules 2001, 34, 5644. (4) Biesalski, M.; Ruhe, J. Macromolecules 1999, 32, 2309. (5) Marutani, E.; Yamamoto, S.; Ninjbadgar, T.; Tsujii, Y.; Fukuda, T.; Takano, M. Polymer 2004, 45, 2231. (6) Xiao, D. Q.; Wirth, M. J. Macromolecules 2002, 35, 2919. (7) Carrot, G.; Diamanti, S.; Manuszak, M.; Charleux, B.; Vairon, I. P. J. Polym. Sci. Part A: Polym. Chem. 2001, 39, 4294. (8) Perruchot, C.; Khan, M. A.; Kamitsi, A.; Armes, S. P.; von Werne, T.; Patten, T. E. Langmuir 2001, 17, 4479. (9) Ejaz, M.; Tsujii, Y.; Fukuda, T. Polymer 2001, 42, 6811. (10) Jeyaprakash, J. D.; Samuel, S.; Dhamodharan, R.; Ruhe, J. Macromol. Rapid Commun. 2002, 23, 277. (11) Tsujii, Y.; Ejaz, M.; Sato, K.; Goto, A.; Fukuda, T. Macromolecules 2001, 34, 8872. (12) Zhou, Q. Y.; Fan, X. W.; Xia, C. J.; Mays, J.; Advincula, R. Chem. Mater. 2001, 13, 2465. (13) Jordan, R.; West, N.; Ulman, A.; Chou, Y. M.; Nuyken, O. Macromolecules 2001, 34, 1606.
zation,15-17 photoinitiated polymerization,18 laser-irradiation induced polymerization,19 block copolymerization,20 and gradient polymerization.21-23 Second, because SIP can be carried out using different polymerization methods, it is compatible with a wide range of synthetic monomers, such as styrene,3 Nisopropylacrylamide,24 acrylamide,6,23 ethylene glycol,25 and methacrylate.8,26,27 Third, SIP is also compatible with substrates of different compositions, shapes, and dimensions: SIP has now been performed on a variety of planar substrates, such as plastics,28 gold,18,29 silica,6 silicon/silicon dioxide,15 and glass,30 and on nanomicroscale structures, such as carbon black,31 clay particles,12 gold nanoparticles,13 multiwall carbon nanotubes,32 magnetite nanoparticles,5 and silica nanoparticles.7 Fourth, SIP methods are also compatible with soft-lithography and dip-pen nanolithography, so that patterned polymer features can be conveniently grown on a surface with a lateral feature size ranging (14) Advincula, R. In Surface-Initiated Polymerization I; Springer: New York, 2006; Vol. 197, p 107. (15) Yoon, K. R.; Chi, Y. S.; Lee, K. B.; Lee, J. K.; Kim, D. J.; Koh, Y. J.; Joo, S. W.; Yun, W. S.; Choi, I. S. J. Mater. Chem. 2003, 13, 2910. (16) Liu, X. G.; Guo, S. W.; Mirkin, C. A. Angew. Chem. Int. Ed. 2003, 42, 4785. (17) Carrot, G.; Rutot-Houze, D.; Pottier, A.; Degee, P.; Hilborn, J.; Dubois, P. Macromolecules 2002, 35, 8400. (18) Schmidt, R.; Zhao, T. F.; Green, J. B.; Dyer, A. J. Langmuir 2002, 18, 1281. (19) Saito, N.; Yamashita, S.; Matsuda, T. J. Polym. Sci. Part A: Polym. Chem. 1997, 35, 747. (20) Wieringa, R. H.; Siesling, E. A.; Werkman, P. J.; Vorenkamp, E. J.; Schouten, A. J. Langmuir 2001, 17, 6491. (21) Xu, C.; Wu, T.; Drain, C. M.; Batteas, J. D.; Fasolka, M. J.; Beers, K. L. Macromolecules 2006, 39, 3359. (22) Xu, C.; Wu, T.; Batteas, J. D.; Drain, C. M.; Beers, K. L.; Fasolka, M. J. Appl. Surf. Sci. 2006, 252, 2529. (23) Wu, T.; Efimenko, K.; Vlcek, P.; Subr, V.; Genzer, J. Macromolecules 2003, 36, 2448. (24) Jones, D. M.; Smith, J. R.; Huck, W. T. S.; Alexander, C. AdV. Mater. 2002, 14, 1130. (25) Ma, H. W.; Hyun, J. H.; Stiller, P.; Chilkoti, A. AdV. Mater. 2004, 16, 338. (26) Mandal, T. K.; Fleming, M. S.; Walt, D. R. Chem. Mat. 2000, 12, 3481. (27) Ejaz, M.; Ohno, K.; Tsujii, Y.; Fukuda, T. Macromolecules 2000, 33, 2870. (28) Ulbricht, M.; Belfort, G. J. Membr. Sci. 1996, 111, 193. (29) Kim, J. B.; Bruening, M. L.; Baker, G. L. J. Am. Chem. Soc. 2000, 122, 7616. (30) Tsubokawa, N.; Satoh, M. J. Appl. Polym. Sci. 1997, 65, 2165. (31) Tsubokawa, N.; Takeda, N.; Kanamaru, A. J. Polym. Sci. Part C: Polym. Lett. 1980, 18, 625. (32) Baskaran, D.; Mays, J. W.; Bratcher, M. S. Angew. Chem., Int. Ed. 2004, 43, 2138.
10.1021/la701630g CCC: $37.00 © 2007 American Chemical Society Published on Web 10/12/2007
Surface-Initiated Enzymatic Polymerization of DNA
from tens of nanometers to the macroscale by patterning the initiator prior to SIP.3,24,33,34 Extending SIP to biological polymers is, however, more problematic. Although solid-phase synthesis methods are widely used to synthesize peptides and oligonucleotides on solid supports, these stepwise approaches have several limitations: (1) they are time-consuming, because the addition of each amino acid or nucleotide requires multiple deprotection, coupling, and protection reactions; (2) only short sequences (peptides oligo(C). The combination of these two sets of data yields the striking observation that the optimal initiator, oligo(A), with the optimal monomer, dATP, yields a poly(A) homopolymer that is ∼50 nm in thickness with a polymerization time of 2 h, which is faster than that of SIP with synthetic monomers (∼50 nm in 12 h)25 and is significantly faster than that of poly(A-T) polymerized by Taq polymerase (∼10 nm in 6 h under the best condition).37 Composition of SIEP-Extended DNA. We further characterized the DNA SAM before and after SIEP by XPS to determine the composition of the ssDNA. Compared to bare gold, overnight incubation with 5′-SH(CH2)6-dTTP25 led to an increase in the sulfur signal (113%) and the appearance of a new phosphorus signal (Table 1). The decrease in the silicon (73%) and gold (22%) signals and the increase in the sulfur signal and the appearance of the phosphorus peak are consistent with the formation of a 5′-SH(CH2)6-dTTP25 SAM on gold. A trace level of sodium was detected on the oligonucleotide SAM on gold, but its level was too low to be reliably quantified. After SIEP for 2 h, there was a substantial decrease in the signal for sulfur (82%), silicon (92%), and gold (92%); an increase in the signal for carbon (20%), oxygen (80%), nitrogen (118%), and phosphorus (217%); and the appearance of a new sodium signal as compared to the oligonucleotide SAM on gold. The new sodium peak and the increase in the phosphorus peak can be attributed to the increased amount of DNA on the surface after SIEP. Despite the deposition of 16.1 nm of DNA, as measured by ellipsometry (38) Petrovykh, D. Y.; Kimura-Suda, H.; Tarlov, M. J.; Whitman, L. J. Langmuir 2004, 20, 429. (39) Poon, K.; Macgregor, R. B. J. Biopolymers 1998, 45, 427.
for SIEP for 2 h, a residual gold signal of ∼2% was observed, which is inconsistent with the ∼10 nm penetration depth of XPS.40 We believe that the residual gold signal must arise from the fact that SIEP is not homogeneous over the area analyzed by XPS, so that bare patches remain where the gold is within the sampling depth of XPS. As a negative control, we performed SIEP with ddTTP, a chain terminator of SIEP, in order to deconvolute the compositional changes caused by the adsorption of the different reagents to the surface from that of the polymerized DNA homopolymers. There were only small differences in signal for carbon (7%) and nitrogen (11%) between SIEP-extended DNA with ddTTP and dTTP. However, the signals for oxygen, phosphorus, sulfur, sodium, silicon, and gold of the negative control and those of DNA SAM after SIEP with dTTP differed by at least 1.7-fold. These results were similar to those obtained for DNA SAM incubated only with TdTase (except for sodium and silicon, for which signals were very low), indicating that the changes in signal for carbon and oxygen and almost half of the change in signal for gold were due to the adsorption of TdTase. The calculation of atomic ratios of carbon, oxygen, nitrogen, and phosphorus provides a clearer view of the changes in composition of DNA SAM before and after SIEP with dTTP. The atomic ratios of P/N, O/N, and C/N for DNA SAM changed from 0.27, 3.52, and 10.27 to 0.40, 2.90, and 5.61, respectively. After SIEP with the ddTTP control, the P/N, O/N, and C/N ratios were 0.06, 1.49, and 5.37, respectively. For oligonucleotides of dTTP, the expected atomic ratios of P/N, O/N, and C/N are 0.5, 3.5, and 5, respectively. Because adsorption of TdTase contributes to the XPS signal for carbon and nitrogen, direct comparison with theoretical estimates of these ratios is not possible. However, the significant increase in the P/N ratio from 0.27 for the oligonucleotide SAM to 0.40 after SIEP with dTTP is consistent with the formation of a DNA overlayer by SIEP. Antibody Binding of SIEP-Extended DNA. To further verify the extension of the DNA homopolymers by SIEP, we synthesized poly(dTTP) from a 5′-SH(CH2)6-dTTP25 SAM on gold and used SPR to measure the binding of the DNA polymer to an antibody that has a binding epitope against (dTTP)6. SPR showed that the TdTase-catalyzed polymerization on the pure SAM of 5′-SH(CH2)6-dTTP25 resulted in a ∼10-fold greater signal than that on a 1:1 mixed SAM of 5′-SH(CH2)6-dTTP25 and SH(CH2)11-EG3OH (Figure 4a). This result is inconsistent with the ellipsometry data, which showed only a 45% difference between these two conditions. This discrepancy may be caused by the fact that SPR measures the changes in the optical properties of a solvated DNA chain, while the ellipsometry measurements were carried out on dried films in air. The origin of this discrepancy is currently under investigation. We found that antibody binding to poly(dTTP)n as measured by SPR strongly depends on the surface density of the oligonucleotide initiator. The antibody binding to the mixed SAM of 5′-SH(CH2)6-dTTP25 was significantly different, compared to (40) Chester, M. J.; Jach, T.; Thurgate, S. J. Vac. Sci. Technol., B 1993, 11, 1609.
11716 Langmuir, Vol. 23, No. 23, 2007
Figure 4. SPR sensorgrams for anti-ssDNA antibody binding with TdTase-extended poly(dTTP) on a pure SAM of 5′-SH(CH2)6-dTTP25 and a mixed SAM of 5′-SH(CH2)6-dTTP25 and SH(CH2)11-EG3-OH in a 1:1 molar ratio (A). Poly-dTTP was synthesized on SAMs by incubating with TdTase, dTTP, and cobalt-containing buffer for 50 min. Twenty-five microliters of anti-poly(dTTP) antibody was subsequently injected at two concentrations (10 and 100 µg/mL) at a flow rate of 1 µL/min. Differences in SPR signals after each step [DNA extension (SIEP), Ab (10 µg/mL), and Ab (100 µg/mL)] are shown in part B. For a negative control, antibodies at two concentrations were injected onto oligonucleotide SAMs that had not been extended by TdTase (no SIEP).
its homogeneous counterpart (Figure 4b). Although the degree of polymerization on the mixed SAM was less than that of the pure SAM, the amount of antibody bound to the DNA was significantly greater. This phenomenon was observed for both antibody concentrations tested, though the effect at 10 µg/mL was greater than that at 100 µg/mL. At 10 µg/mL of antibody, the amount of antibody bound to the DNA polymerized on the mixed SAM was ∼40 times the negative control, while that on the pure SAM was only 20% of the negative control (Figure 4b). This result highlights the importance of the packing density of extended DNA on antibody binding, where mixed SAMs with a lower density of extended DNA resulted in more antibody binding, possibly due to greater steric access to the DNA polymer chains. SIEP on Micropatterned DNA SAMs. SIEP is also compatible with commonly used micropatterning techniques for generating DNA micro- and nanostructures. Figure 5 shows the formation of micropatterned poly(dTTP) features that were grown from a 5′-SH(CH2)6-dTTP25 SAM that was micropatterned by
Chow and Chilkoti
Figure 5. Images of SIEP-extended DNA on a template of DNA SAM (5′-SH(CH2)6-dTTP25) micropatterned using microchannels. AFM tapping-mode images (A, side view; B, top view) and a line profile of the area highlighted with arrows (C).
microfluidics. The height of the poly(dTTP) microstructures grown by SIEP, as measured by AFM in air, was ∼60 nm. When compared to the thickness of extended DNA measured using ellipsometry, the height measured by AFM was somewhat larger, probably due to the variation in surface density of oligonucleotide initiators caused by different DNA SAM preparation methods. This result demonstrates that DNA microstructures can be conveniently synthesized by combining SIEP with soft lithography. Control experiments in which the oligonucleotide thiol initiator was omitted during microfluidic patterning did not lead to any extension of DNA (data not shown).
Discussion TdTase is an enzyme expressed in immune cells and leukemia/ lymphoma cells41 and was first isolated in 1965.42 It plays an important role in generating antibody diversity by the insertion of extra nucleotides between the V and D gene segments of the H chain gene.41 Although TdTase has also been used as a reagent in molecular biology for end-labeling,42 cloning by reverse amplification of chain ends (RACE),43 and detection assays for apoptosis (TUNEL assay),44 its use in surface biochemical reactions had not been explored until our recent paper35 and the (41) Eun, H.-M. In Enzymology Primer for Recombinant DNA Technology; Eun, H.-M., Ed.; Academic Press: New York, 1996; p 477. (42) Yoneda, M.; Bollum, F. J. J. Biol. Chem. 1965, 240, 3385. (43) Sambrook, J.; Russell, D. W. In Molecular Cloning: A Laboratory Manual; Sambrook, J., Russell, D. W., Eds.; Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY, 2001; p 54. (44) Gavrieli, Y.; Sherman, Y.; Ben-Sasson, S. A. J. Cell Biol. 1992, 119, 493.
Surface-Initiated Enzymatic Polymerization of DNA
subsequent papers by Ruysschaert et al. and Anne et al.36,45 In the light of this lacuna and the utility of this reaction for surfaceinitiated polymerization and biochemical amplification, we examined SIEP by TdTase in detail in this paper. Using ellipsometry and XPS, we verified the feasibility of rapidly generating long DNA chains from an oligonucleotide SAM by SIEP. We show that the extension of DNA requires surface-bound oligonucleotide initiator and that the amount of polymerized DNA depends on the surface density of oligonucleotide initiators, analogous to the chemical SIP of synthetic monomers.25 In addition, the degree of extension also largely depends on the type of mononucleotides and oligonucleotide initiators, which is similar to the behavior of TdTase-catalyzed polymerization of an oligonucleotide in solution.42 In general, the nucleotide preference of TdTase in SIEP is dATP > dTTP . dGTP ≈ dCTP, regardless of the type of oligonucleotide initiators used. An important parameter, however, that has not been explored in the current study is the length of the oligonucleotide initiator, which has been shown to play a role in controlling the initial rate of polymerization in solution.42 As shown by the SPR experiments, the SIEP-extended DNA is capable of binding to an antibody that is specific to singlestranded DNA. The mole fraction of DNA-SH in solution used to create the mixed SAM has a large effect on the amount of antibody binding, indicating that the surface density of the initiator controls polymer density, which in turn controls the accessibility of the ssDNA chains by the antibody. Finally, we show that SIEP can be combined with soft lithography to fabricate microstructures of DNA. SIEP can also be combined with other lithographic techniques such as edge-effect lithography, e-beam lithography,35 and patterning using microwells (Wellpat).46 A greater level of functional sophistication can also be embedded (45) Anne, A.; Bonnaudat, C.; Demaille, C.; Wang, K. J. Am. Chem. Soc. 2007, 129, 2734.
Langmuir, Vol. 23, No. 23, 2007 11717
into these structures by sequentially building block copolymers with different molecular recognition properties using different natural and unnatural nucleotides. We believe that these mesoand microscale structures will be useful as intelligent molecular scaffolds to dock nanoparticles and biomolecules for the fabrication of hybrid structures at the meso-microscale and for the development of new supramolecular architectures for biosensors and clinical diagnostics.
Conclusions We have presented a method to synthesize polynucleotide chains on solid surfaces using SIEP with TdTase. Ellipsometry showed a substantial increase in the thickness of DNA SAM upon SIEP with TdTase and that the choice of oligonucleotide initiators strongly influenced the thickness of the DNA polymer. In addition, the choice of nucleotide monomers also had a significant effect on the extent of polymerization, with dATP > dTTP . dGTP ≈ dCTP. The surface density of oligonucleotide initiators affected not only the overall extent of polymerization but also the packing density of the extended DNA, which in turn affected the ability of an ssDNA-binding antibody to bind to the polymerized DNA. Finally, we demonstrated that SIEP with TdTase is compatible with micropatterning techniques, which provides new opportunities for the fabrication of micro- and nanostructures with controlled sequence, length, and molecular recognition properties. Acknowledgment. This research was financially supported by the National Science Foundation (NSF) through a Nanoscale Interdisciplinary Research Teams (NIRT) grant (NSF-0609265). LA701630G (46) Hyun, J.; Chilkoti, A. J. Am. Chem. Soc. 2001, 123, 6943.