Langmuir 2008, 24, 927-931
927
Controlling the Reactivity of Adsorbed DNA on Template Surfaces R. Kaufmann, I. Averbukh, and R. Naaman* Department of Chemical Physics, Weizmann Institute, RehoVot 76100, Israel
S. S. Daube Chemical Research Support, Weizmann Institute, RehoVot 76100, Israel ReceiVed September 10, 2007. In Final Form: NoVember 9, 2007 The ability to place DNA on surfaces with increased and controllable reactivity is of fundamental importance in the development of next-generation DNA and protein biochips. The present work demonstrates the ability to control both the localization of the DNA on a surface and its reactivity by a self-assembly approach that is dependent on two variables: DNA structure and surface environment. Here we utilize a two-step adsorption scheme to control the adsorption and reactivity of DNA embedded within two types of alkyl thiol monolayers (either methyl-terminated or hydroxyl-terminated). In addition, by changing the structure of the chemisorbed DNA from fully single stranded to a 50% double stranded at its side adjacent to the surface, we were able to observe a clear dependence of DNA reactivity on both the DNA structure and the type of alkyl thiol monolayer covering the surface. The adsorption and the reactivity yield of the DNA were monitored either by its ability to hybridize to a complementary target molecule or by an enzymatic reaction involving DNA phosphorylation catalyzed by the enzyme T4 polynucleotide kinase.
Introduction The ability to bind DNA molecules to surfaces in the form of a monolayer is widely used in the construction of DNA microarrays. DNA adsorption on surfaces has therefore been characterized extensively by various surface techniques.1-8 Numerous studies have also demonstrated the ability to perform enzymatic manipulations on adsorbed DNA, such as PCR, DNA ligation, and restriction enzyme digestion.9-14 Improved performance of micro arrays and sensors is expected to occur by reducing their size concomitantly with an increase in the number of samples processed simultaneously. Therefore, understanding the structural and environmental influences on the adsorbed biomolecules is essential for controlling their position and reactivity, which is of fundamental importance in the development of DNA and protein biochips and nanodevices.15 Immobilizing DNA to a gold surface is commonly performed by direct adsorption of thiolated DNA oligomers onto bare gold, (1) Strother, T.; Cai, W.; Zhao, X. S.; Hamers, R. J.; Smith, L. M. J. Am. Chem. Soc. 2000, 122, 1205. (2) Singhal, P.; Kuhr, W. G. Anal. Chem. 1997, 69, 4828. (3) Steel, A. B.; Herne, T. M.; Tarlov, M. J. Anal. Chem. 1998, 70, 4670. (4) Herne, T. M.; Tarlov, M. J. J. Am. Chem. Soc. 1997, 119, 8916. (5) Csa´ki, A.; Mo¨ller, R.; Straube, W.; Ko¨hler, J. M.; Fritzsche, W. Nucleic Acids Res. 2001, 29, e81. (6) Caruso, F.; Rodda, E.; Furlong, D. N. Anal. Chem. 1997, 69, 2043. (7) Demers, L. M.; Mirkin, C. A.; Mucic, R. C.; Reynolds, R. A., III; Letsinger, R. L.; Elghanian, R.; Viswanadham, G. Anal. Chem. 2000, 72, 5535. (8) Mourougou-Candoni, N.; Naud, C.; Thibaudau, F. Langmuir 2003, 19, 682. (9) Gerry, N. P.; Witowski, N. E.; Day, J.; Hammer, R. P.; Barany, G.; Barany, F. J. Mol. Biol. 1999, 292, 251. (10) Kim, J. H.; Hong, J-A.; Yoon, M.; Yoon, M. Y.; Jeong, H-S.; Hwang, H. J. J. Biotechnol. 2002, 96, 213. (11) Pen˜a, S. R. N.; Raina, S.; Goodrich, G. P.; Fedroff, N. V.; Keating, C. D. J. Am. Chem. Soc. 2002, 124, 7314. (12) Frutos, A. G.; Smith, L. M.; Corn, R. M. J. Am. Chem. Soc. 1998, 120, 10277. (13) Bamdad, C. Biophys. J. 1998, 75, 1997. (14) O’Brien, J. C.; Stickney, J. T.; Porter, M. D. J. Am. Chem. Soc. 2000, 122, 5004. (15) Gillmor, S. D.; Thiel, A. J.; Strother T. C.; Smith, L. M.; Lagally M. G. Langmuir 2000, 16, 7223.
followed by a short exposure of these DNA-modified gold surfaces to alkyl thiols. The alkyl thiols compete with the nonspecific interaction of the DNA bases with the surface, resulting in a more extended conformation of the DNA.4 Recently we demonstrated an alternative approach in which thiolated DNA is adsorbed into the defects within a monolayer of alkyl thiolmodified gold, rather than on bare gold.16 We showed that this methodology is advantageous for adsorption of long double stranded (ds) DNA molecules, by reducing the amount of nonspecific adsorption. In addition, we observed a dramatic increase in the reactivity of short dsDNA embedded within the defects in an enzymatic ligation reaction, as compared to its reactivity on bare gold. In the present work we demonstrate that it is indeed possible to control both the localization of the DNA on a surface and its reactivity by a self-assembly approach. By prepatterning a gold surface with two types of alkyl-thiol monolayers, a solution of DNA molecules can be added to the entire surface. The adsorption efficiency of the DNA molecules depends on the preadsorbed species and therefore a pattern of DNA can be obtained. In addition, the DNA molecules in each area would have different reactivity depending on their interaction with the alkyl thiol in their vicinity (Figure 1A) and on their structure. Experimental Section Substrate Preparation. The adsorption was performed on goldcoated silicon (111) wafers. The gold-coated wafer was incubated in 0.1 mM alkylthiol solution for 10 min at room temperature in order to produce a nonperfect monolayer. Three types of monolayers were produced: octadecanethiol (hydrophobic), dodecanethiol (hydrophobic), and a mercaptoundecanol (hydrophilic). The layer was characterized by contact angle and ellipsometric measurements (Supporting Information). DNA Preparation. All probes were based on a 28nt long DNA oligomer containing a C3 thiol linker (MWG) at its 3′ end. The 28nt probe was hybridized with a 14nt complementary DNA oligomer by mixing 10 µM of each oligomer in 25 mM Tris-HCl pH 7.5, 0.2M NaCl buffer (buffer A). The mixture was incubated at 90 °C for 10
10.1021/la702799v CCC: $40.75 © 2008 American Chemical Society Published on Web 01/10/2008
928 Langmuir, Vol. 24, No. 3, 2008
Kaufmann et al.
Figure 1. (a) Proposed conformations of ssDNA and dsDNA within a methyl-terminated (I) and a hydroxyl-terminated (II) monolayer. In (I), the single strand region of the rigid dsDNA probe is directed toward the solution while in the flexible ssDNA probe it is in proximity to the monolayer. Bases embedded within the methyl-terminated monolayer are directed toward the layer and are therefore not available for interaction with the complementary target. In (II), the bases embedded within a hydroxyl-terminated monolayer are randomly oriented and can therefore interact with the complementary target in the solution. (b) Illustration of the sequence and structure of the DNA probes and DNA target. The complementary sequences of the probe and target are in blue and red respectively. The probes are bound to a gold surface (yellow) covered with an organic monolayer (green). min, after which the reaction was cooled down slowly to room temperature. The DNA was radioactively labeled by the enzyme T4 poly nucleotide kinase (New England Biolab) and γ-P32-ATP (3000 Ci/mmol, NEN). After removing any unincorporated γ-P32-ATP by chromatography on a Micro Spin 6 column (Bio Rad), the radiolabeled oligomers were combined with unlabeled oligomers in a ratio of 1:4 to a final concentration of 10 µM.
To perform parallel reactions, an RTV thin film containing 4 mm diameter holes was attached to the gold surface producing small RTV reaction wells (Supporting Information). For DNA adsorption in between the nonperfect monolayers,16 12.5 µL of 5 µM from the different probe oligomers (in buffer A) (16) Aqua, T.; Naaman, R.; Daube, S. Langmuir 2003, 19, 10573.
ReactiVity of Adsorbed DNA on Template Surfaces
Figure 2. Side view of the experimental procedure. The RTV layer is shown in pink while the other color codes are as in Figure 1b. were applied into the wells. After filling the wells, they were sealed with a thin elastic RTV sheet. The entire setup was then incubated for overnight adsorption at room temperature. The RTV cover was removed and the wells were emptied and rinsed with buffer A to wash away excess DNA. This process is described schematically in Figure 2A-C. The amount of probe that had remained attached to the surface was quantified using a phosphorimager (FUJI). (See Supporting Information for further details.) DNA Reactivity Assays. For the hybridization experiment, 12.5 µL of 1 µM radioactive labeled complementary target DNA solution (in buffer A) was applied into the wells. The wells were sealed again with the thin elastic RTV sheet and the entire setup was incubated at 50 °C for a few hours. To end the incubation, the wells were opened and emptied, followed by removal of the RTV array and washing of the slide (Figure 2D,E). The amount of complementary target that had remained attached to the probes was quantified using a phosphorimager (FUJI). (See Supporting Information for further details.)
Results and Discussion The effect of prepatterning the substrate with either methylterminated or hydroxyl-terminated monolayer on the adsorption and reactivity of adsorbed DNA was demonstrated by using two types of DNA oligomers (probes): a 28 nucleotide (nt) long fully single strand DNA (referred to as ssDNA) and a 14 base pairs (bp) long double strand DNA with a 14nt long single strand DNA extension (referred to as dsDNA) (Figure 1B). The DNA
Langmuir, Vol. 24, No. 3, 2008 929
probes were bound to a gold surface coated with either mercaptoundecanol, octadecanthiol, or dodecanethiol molecules. Once adsorbed, the reactivity of these two DNA probes was evaluated by adding radioactively labeled complementary DNA (target) to non-labeled adsorbed probes and calculating the efficiency of hybridization. All of the experiments were therefore performed in three steps: (i) Preparing surfaces covered with self-assembled monolayers (SAM) of either methyl-terminated or hydroxyl-terminated alkyl thiols. (Figure 2A). (ii) Adsorbing of the ssDNA and dsDNA probes in between the SAM molecules. (Figure 2B,C). (iii) Reacting the adsorbed probe DNA with either the target DNA (hybridization reaction, Figure 2D,E) or with the enzyme T4 polynucleotide kinase (5′-end DNA phosphorylation). As can be seen in Figure 3, the two types of probes (ssDNA and dsDNA) adsorbed with different yields on the methylterminated surface, resulting in a much higher coverage of ssDNA than dsDNA. This effect was reversed when the two types of DNA were adsorbed on the hydroxyl-terminated surface. In addition to the selectivity in the adsorption of DNA, we also observed selectivity in the reactivity of the DNA probes once on the surface (Figure 4). Although the dsDNA probe did not adsorb efficiently onto the methyl-terminated surface, those molecules that did adsorb were highly available to hybridization. The opposite trend was observed for the ssDNA probe. That is, the ssDNA probes adsorbed with high efficiency onto the methyl-terminated surface, but were not available to hybridization. The picture was reversed for a hydroxyl-terminated surface: the ssDNA probe did not adsorb efficiently within the hydroxyl-terminated monolayer, but its reactivity was high, while the dsDNA probe adsorbed with higher efficiency than the ssDNA probe but demonstrated reduced hybridization efficiency. These results are summarized quantitatively in Figure 4A showing the reactivity level of each of the probes adsorbed within the two types of SAMs as expressed by the ratio of target molecules to probe molecules per surface area. These ratios remain the same even when the reactions were left overnight (data not shown) suggesting that the differences are not due to kinetic effects. Replacing the octadecanthiol with dodecanethiol monolayer resulted in similar results suggesting that the differences observed between the octadecanthiol and the mercapto-undecanol were due to the end group and not the monolayer thickness. The specificity of both the adsorption and the reactivity were verified by showing that both a non-thiolated DNA probe (data not shown) and a DNA probe with no complementary sequences to the DNA target (labeled as non, Figure 4A) bound to the surface at very low efficiencies. In addition, we showed that the increase in DNA reactivity was not limited to hybridization, but was more general, by performing an enzymatic reaction on the surface. As can be seen in Figure 4B, similarly to hybridization, the dsDNA probe was phosphorylated with a radioactive phosphate much more efficiently than the ssDNA probe by the enzyme T4 polynucleotide kinase. To demonstrate the utilization of surface patterning to control DNA reactivity in a self-assembly manner, we sought to show that the rules formulated above can be applied to heterogeneous surfaces that had been prepatterned with methyl and hydroxylterminated alkyl thiols on adjacent regions. Dipping such a slide into a radioactively labeled DNA probe (either ssDNA or dsDNA) resulted in separation of the DNA molecules (Figure 3A). The ssDNA and dsDNA probes segregated on the heterogeneous surface according to the variable interaction of the DNA with
930 Langmuir, Vol. 24, No. 3, 2008
Kaufmann et al.
Figure 3. (a) Phosphorimaging scan of a surface patterned with organic SAMs and DNA. The upper part of each slide was modified with a hydroxyl-terminated SAM while the lower part with a methyl-terminated one. The probe and target bars refer to radioactive-labeled probe DNA and target DNA, respectively. (b) DNA densities (molecules/mm2) of adsorbed ssDNA (white), dsDNA (lined) probes, and their hybridized targets (gray) on the methyl-terminated (front) and hydroxyl-terminated (back) SAMs. The accuracy of the results is (10%.
each of the SAMs. Hybridization of target DNA molecules applied to the entire slide was also heterogeneous according to the local environment. This control over the local adsorption and reactivity of DNA on a surface is shown quantitatively in Figure 3B. The change in reactivity observed cannot be attributed simply to differences in DNA density on the two SAMs, since in both cases the density of DNA is far below a full monolayer, yet the DNA has been found to be distributed quite homogeneously within the thiolated monolayers.17 Although various models may explain our observations, in what follows we propose a simple model that is consistent with previous studies.18 The reactivity of the DNA adsorbed within the methyl-terminated monolayer can be explained by the interaction of the aromatic bases of the ssDNA with the hydrophobic molecules of the monolayer. Figure 1a (left) presents schematically the proposed structures of ssDNA and dsDNA when adsorbed within such a methyl-terminated SAM that has a hydrophobic nature. The 28nt long oligomers are longer than the thickness of the monolayer. Hence, while the stiff dsDNA stands perpendicular to the surface, the ssDNA (17) Nogues, C.; Cohen, S. R.; Daube, S. S.; Naaman, R. Phys. Chem. Chem. Phys. 2004, 6, 4459. (18) Matsuda, S.; Romesberg, F. E. J. Am. Chem. Soc. 2004, 126, 14419.
tends to lie on the monolayer. As a result, its bases are interacting with the alkyl chain, while the charged phosphates are exposed toward the water. Because the bases are not exposed to the solution, the reactivity of the ssDNA would be hindered. This conclusion is supported by a sequence dependence study (data not shown) in which the sequence of the probe oligomer was changed from that shown in Figure 1b to homopolymers of either all-adenine (oligoA) or an all-cytosine (oligoC) DNA. The oligoA adsorbed with high yield to the methyl-terminated surface but reacted poorly with its complementary target (0.1 ( 0.05). The less hydrophobic oligoC, adsorbed with a lower surface coverage to the methyl-terminated surface but was much more reactive (0.6 ( 0.1). This experiment verified that the stronger is the interaction of the bases with the methyl-terminated monolayer, the lower is the reactivity of the DNA. We ruled out the possibility that this result stemmed from the higher stability of G/C base pairs, by reversing the interaction and reactivity on a hydroxylterminated surface. The situation is very different with a hydroxyl-terminated monolayer (Figure 1a right). In this case, the interaction of the polar groups of the DNA with the polar groups of the monolayer is comparable to the interaction water-DNA. As a result, some of the bases are exposed toward the solution, enabling the DNA to react. In addition, the results obtained at the hydroxyl-
ReactiVity of Adsorbed DNA on Template Surfaces
Langmuir, Vol. 24, No. 3, 2008 931
Namely, on the hydroxyl-terminated surface the density of ssDNA and dsDNA probes was very similar yet they reacted quite differently. The present study provides a new avenue for controlling DNA reactivity when adsorbed on surfaces, by showing how the conformation of the DNA attached to a surface can be controlled by both the local environment of the surface and the structure of the DNA. It is the combination of these two factors that would eventually determine the extent of DNA reactivity. The present study therefore reveals new parameters that can be used to tune the reactivity of DNA when adsorbed on surfaces, but also provides insights into molecular genetics,19,20 biosensor design,21,22 DNA biophysics,23-25 and basic separation theory.26 Acknowledgment. This research was partially supported by the Grand Center for Sensors and Security. Supporting Information Available: Experimental procedures. Schematic of thin elastic RTV wells on a gold slide covered with an organic monolayer and RTV small reaction wells. This material is available free of charge via the Internet at http://pubs.acs.org. LA702799V
Figure 4. (A) DNA hybridization efficiency on octadecanthiol (black), dodecanethiol (gray), and mercapto-undecanol (white) monolayers obtained for dsDNA (ds), ssDNA (ss) and noncomplementary to target (non) probes. Hybridization efficiency is calculated as the number of moles of target DNA divided by the number of probe DNA for a given surface area. (B) Enzymatic phosphorylation efficiency on octadecanthiol monolayer for dsDNA (ds), ssDNA (ss), and a control reaction without enzyme (non).
terminated surface enabled us to rule out the possibility that the variable reactivity of the DNA was due to its density.
(19) Herrick, J.; Michalet, X.; Conti, C.; Schurra, C.; Bensimon, A. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 222. (20) Lyubchenko, Y. L.; Shlyakhtenko, L. S. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 496. (21) Chan, V.; McKenzie, S. E.; Surrey, S.; Fortina, P.; Graves, D. J. J. Colloid Interface Sci. 1998, 203, 197. (22) Chan, V.; Graves, D. J.; Fortina, P.; McKenzie, S. E. Langmuir 1997, 13, 320. (23) Bensimon, A.; Simon, A.; Chiffaudel, A.; Croquette, V.; Heslot, F.; Bensimon, D. Science 1994, 265, 2096. (24) Xue, Q.; Yeung, E. S. Nature 1995, 373, 681. (25) Fan, F.-R. F.; Bard, A. J. Science 1995, 267, 871. (26) Kang, S. H.; Shortreed, M. R.; Yeung, E. S. Anal. Chem. 2001, 73, 1091 and references cited therein.
