Langmuir 2007, 23, 8607-8613
8607
Transferring Complementary Target DNA from Aqueous Solutions onto Solid Surfaces by Using Affinity Microcontact Printing Hua Tan, Shisheng Huang, and Kun-Lin Yang* Department of Chemical and Biomolecular Engineering, National UniVersity of Singapore, 4 Engineering DriVe 4, Singapore 117576 ReceiVed February 9, 2007. In Final Form: May 8, 2007 In this paper, we report a method of transferring complementary target DNA from an aqueous solution onto a solid surface by using affinity microcontact printing. In this approach, the probe DNA is first immobilized on the surface of an aminated poly(dimethylsiloxane) (PDMS) stamp. After a complementary target DNA hybridizes with the probe DNA on the stamp surface, the PDMS stamp is printed on an aminated glass slide. By using fluorescent microscopy, we show that only complementary target DNA, but not noncomplementary DNA, can be captured onto the surface of the stamp and then transferred to the aminated glass slide. The transfer of DNA can be attributed to the stronger electrostatic attraction between DNA and amine groups compared to the hydrogen bonds between the hybridized DNA molecules. We also investigate several factors that may influence the transfer of DNA, such as the surface density of amine groups, hybridization conditions, and contamination from unreacted PDMS monomers.
Introduction Affinity microcontact printing (RCP) was first developed by Bernard and co-workers1 several years ago. In their work, the surface of a poly(dimethylsiloxane) (PDMS) stamp was functionalized with anti-mouse IgG which selectively captured 125I-labeled mouse IgG in a crude biological sample. After rinsing to remove unbound molecules from the stamp surface, the stamp was brought into conformal contact with a solid surface. Because of the stronger interaction between the surface and proteins, the captured molecules were transferred to the solid surface. The same research group also demonstrated that protein microarrays can be fabricated by using RCP.2 In addition, Abbott et al. employed RCP to transfer anti-biotin IgG or epidermal growth factor receptor (EGFR) to surfaces from affinity stamps and imaged them using liquid crystals.3-5 Recently, Lange et al. and Thibault et al.6,7 have extended the microcontact printing (µCP) technique to the transfer of DNA onto solid surfaces for building DNA microarrays. In both studies, single-stranded DNA was first captured on the surfaces of PDMS stamps with either electrostatic attraction or hydrophobic interactions between the DNA and the stamp surfaces. The stamps were then printed onto a substrate with amine functional groups on its surface. However, in both studies, only single-stranded DNA was used in the experiment; therefore, it is not clear whether this technique can be used to transfer double-stranded DNA. More recently, Crooks et al.8,9 reported a method for transferring biotin-functionalized * To whom correspondence should be addressed. E-mail: cheyk@ nus.edu.sg. (1) Bernard, A.; Fitzli, D.; Sonderegger, P.; Delamarche, E.; Michel, B.; Bosshard, H. R.; Biebuyck, H. Nat. Biotechnol. 2001, 19, 866-869. (2) Renault, J. P.; Bernard, A.; Juncker, D.; Michel, B.; Bosshard, H. R.; Delamarche, E. Angew. Chem., Int. Ed. 2002, 41, 2320-2323. (3) Jang, C. H.; Tingey, M. L.; Korpi, N. L.; Wiepz, G. J.; Schiller, J. H.; Bertics, P. J.; Abbott, N. L. J. Am. Chem. Soc. 2005, 127, 8912-8913. (4) Tingey, M. L.; Snodgrass, E. J.; Abbott, N. L. AdV. Mater. 2004, 16, 1331-1336. (5) Tingey, M. L.; Wilyana, S.; Snodgrass, E. J.; Abbott, N. L. Langmuir 2004, 20, 6818-6826. (6) Lange, S. A.; Benes, V.; Kern, D. P.; Horber, J. K. H.; Bernard, A. Anal. Chem. 2004, 76, 1641-1647. (7) Thibault, C.; Berre, V. L.; Casimirius, S.; Trevisiol, E.; Francois, J.; Vieu, C. J. Nanobiotechnol. 2005, 3, 1-12. (8) Lin, H. H.; Sun, L.; Crooks, R. M. J. Am. Chem. Soc. 2005, 127, 1121011211.
hybridized DNA from PDMS stamps to streptavidin-modified surfaces. The transfer mechanism was based on the affinity between biotin and streptavidin. Thus, these DNA molecules need to be labeled with biotin first, which limits the applicability of this method. In this study, we propose a new method of transferring hybridized DNA from the stamp surface onto an aminated solid surface by using the electrostatic attraction between the positively charged amino groups and the negatively charged DNA molecules. The transfer mechanism utilized in this study is viable because the electrostatic attraction between the positive amino groups and the negative DNA molecules is generally stronger than the hydrogen bonds between the probe DNA and the target DNA. A schematic illustration of this contact printing process is shown in Figure 1. In the first step, the surface of a PDMS stamp is functionalized with a monolayer of molecules containing primary amines, and then the probe DNA is immobilized on the aminated surface of the PDMS stamp. Next, the probe DNA on the PDMS stamp hybridizes with the complementary target DNA, and the stamp is printed onto an amine-terminated glass slide. Finally, the complementary target DNA is transferred from the PDMS stamp to an aminated glass slide if the electrostatic force between the positive amine groups and the negative DNA molecules is stronger than the hydrogen bonds between the probe DNA and target DNA. To ensure that only the target DNA is transferred during the contact printing process, it is important to prepare a robust stamp surface with immobilized DNA probes which do not detach from the stamp surface together with the target DNA upon contact printing. Unfortunately, several issues associated with microcontact printing have been reported in the past.10 First, unreacted PDMS monomers may transfer from the PDMS stamp to the solid surface during the printing process, which may result in false signals of DNA transfer. Second, since the hydrophobic surface of the PDMS stamp is not suitable for linking DNA, the stamp surface needs to be treated with oxygen plasma and then modified with organosilanes to introduce proper linker groups onto the surface. However, organosilanes tend to form multilayers, (9) Lin, H. H.; Kim, J.; Sun, L.; Crooks, R. M. J. Am. Chem. Soc. 2006, 128, 3268-3272. (10) Quist, A. P.; Pavlovic, E.; Oscarsson, S. Anal. Bioanal. Chem. 2005, 381, 591-600.
10.1021/la701258c CCC: $37.00 © 2007 American Chemical Society Published on Web 06/26/2007
8608 Langmuir, Vol. 23, No. 16, 2007
Figure 1. Schematic illustration of the transfer of complementary DNA from an aqueous solution onto a solid surface by using affinity microcontact printing.
which can hydrolyze upon contact with aqueous solutions. Third, since DNA can adsorb nonspecifically on the surface, we need to take measures to prevent this nonspecific adsorption. In this paper, we will address these issues by properly designing the contact printing processes. In addition, we will also study other influencing factors for the transfer of DNA, such as the surface density of amine groups and the surface density of probe DNA. Experimental Section Materials. All glass slides were purchased from Marienfeld (Germany). BSA (bovine serum albumin), SDS (sodium dodecyl sulfate), APES (3-aminopropyltriethoxysilane), and AEAPS (N(3-(trimethoxysilyl)propyl)-ethylenediamine) were purchased from Sigma-Aldrich (Singapore). Methanol (HPLC grade) was purchased from Merck (Darmstadt, Germany). Sodium chloride sodium citrate buffer (SSC; 1×), containing 0.150 M sodium chloride and 0.015 M sodium citrate, was diluted from 20× SSC (1st BASE, Singapore). DNA 20-mers used in this work, including 5′-TCAGGTTTAGTACCAGAACA-3′ (D1), 5′-TGTTCTGGTACTAAACCTGA-3′FAM (D2), and 5′-ATGCATGCATGCATGCATGC-3′-FAM (D3), were obtained from 1st BASE (Singapore). DNA 25-mers, including 5′-CTGCATGTTCTGGTACTAAACCTGA-3′ (D4), FAM-5′TCAGGTTTAGTACCAAGACATGCAG-3′ (D5), and FAM-5′GCTTTTGCATATTATATCGAGCCAC-3′ (D6), were obtained from Research Biolabs (Singapore). All aqueous solutions were prepared in deionized water with a resistance of 18.2 MΩ. The PDMS stamps were prepared from Sylgard 184 (Dow Corning, Midland, MI). Cleaning of Substrates. The silicon wafers were cleaned by using Piranha solution (70% H2SO4/30% H2O2) at 80 °C for 1 h, followed by rinsing with copious amount of deionized water and ethanol and then dried under a stream of nitrogen. (Warning! Piranha solution should be handled with extreme caution; in some circumstances, most probably when it is mixed with significant
Tan et al. quantities of an oxidizable organic material, it can detonate unexpectedly!) Chemical Modification of Silicon Wafers with APES and AEAPS. To chemically modify a silicon wafer to present amine groups on the surface, we used two amine-bearing organosilanes, AEAPS and APES, for surface modifications. When water was used as the solvent, the silicon wafer was immersed in an aqueous solution containing 5% (v/v) AEAPS or APES at 50 °C. When a mixture of 95% methanol and 5% water was used as the solvent, the silicon wafers were incubated in the solvent containing 1% (v/v) AEAPS or APES at room temperature. After silanization, the silicon wafers were cleaned with the respective solvent and then baked in the oven at 100 °C for 3 h. Immobilization of DNA on Aminated Silicon Wafers and PDMS Stamps. Ten microliters of 10 mM TE buffer containing 100 µM FAM-labeled DNA (D2) was mixed with 10 µL of 6× SSC and 30 µL of 3× SSC. Next, 2 µL of this fleshly prepared DNA solution (with a final concentration of 20 µM) was placed on the surface of an aminated silicon wafer. For immobilization of DNA on a PDMS stamp, 50 µL of the same DNA solution was dispensed onto the surface of the PDMS stamp until the solution covered the entire surface. The silicon wafer or the PDMS stamp was then incubated at 80 °C in a humidified chamber for 3 h to promote cross-linking between the DNA and surface amine groups through either the formation of chemical bonds between thymine and the surface amino groups or strong electrostatic interactions.11-17 After incubation, the wafers or PDMS stamps were first rinsed with 2× SSC containing 0.02% (w/v) SDS for 30 s and then rinsed twice with 1× SSC for 30 s and once with 1× SSC for 2 min. The substrates were then cleaned with 10 mM TE buffer for 30 s, 5 mM TE buffer for 30 s, and finally with 5% (v/v) ethanol in water for 30 s. DNA Hybridization on Silicon Wafers and PDMS Stamps. In SSC Buffer. A substrate decorated with probe DNA (either D1 or D4) was immersed in a blocking solution (0.2 g of BSA in 3 mL of 20× SSC and 16.8 mL of DI water) for 10 min to block the surface. The substrate was then incubated in 10 µL of 4× SSC buffer (with 0.01% (w/v) SDS) containing 50 µM FAM-labeled target DNA (D2 or D3) covered by a lifter slip (Erie Scientific, Portsmouth, NH). The hybridization was carried out at 42 °C for 4 h. After DNA hybridization, the substrates were rinsed following the procedure mentioned above. In TE Buffer. After BSA blocking, the substrate was incubated in 10 µL of TE hybridization buffer (1 M NaCl, 10 mM tris-HCl, 1 mM EDTA, and 0.01% (w/v) SDS, pH ) 7.39) containing 50 µM FAM-labeled target DNA (D2 and D3 for D1 and D5, respectively, and D6 for D4). The hybridization was carried out underneath a lifter slip at room temperature for 12 h. After DNA hybridization, the substrates were also rinsed following the procedure mentioned above. Affinity Microcontact Printing. The PDMS stamps were prepared by casting Sylgard 184 on a silicon master with 400 µm square features fabricated by using photolithography (for affinity microcontact printing) or on a homemade Teflon mold (for ellipsometry experiments). The stamps were degassed in vacuum to remove air bubbles and then cured at 100 °C for 3 h. To clean a PDMS stamp, a Soxhlet device was used to extract the unreacted starting materials of PDMS in ethanol. Finally, the clean PDMS stamp was cured at 100 °C for 1 h to vaporize any ethanol trapped inside the PDMS stamp. The PDMS stamp was then oxidized using oxygen plasma (100 W, 1 min) to form a thin silicon oxide layer on the stamp surface. Subsequently, the oxidized PDMS stamp was (11) Nierzwicki-Bauer, S. A.; Gebhardt, J. S.; Linkkila, L. W.; K. BioTechniques 1990, 9, 472-478. (12) http://www.eriemicroarray.com/support/tech.aspx#proto. (13) Ehrenreich, A. Appl. Microbiol. Biotechnol. 2006, 73, 255-273. (14) Taroncher-Oldenburg, G.; Griner, E. M.; Francis, C. A.; Ward, B. B. Appl. EnViron. Microbiol. 2003, 69, 1159-1171. (15) Wittenberg, A. H. J.; Lee, T.; Cayla, C.; Kilian, A.; Visser, R. G. F.; Schouten, H. J. Mol. Genet. Genomics 2005, 274, 30-39. (16) Gravesen, A.; Kallipolitis, B.; Holmstrom, K.; Hoiby, P. E.; Ramnath, M.; Knochel, S. Appl. EnViron. Microbiol. 2004, 70, 1669-1679. (17) Clarke, B.; Rahman, S. Theor. Appl. Genet. 2005, 110, 1259-1267.
DNA Transfer Using Affinity Microcontact Printing
Langmuir, Vol. 23, No. 16, 2007 8609
Figure 2. Ellipsometric thicknesses (left) and water contact angles (right) on the surfaces of APES- and AEAPS-modified silicon wafers. The silanization processes were performed either in the aqueous phase (a and b) or in the methanol/water mixture (c and d). functionalized with APES by immersion of the stamp in the methanol/ water mixture (95% methanol and 5% water) containing 1% (v/v) Two different organosilanes, APES and AEAPS, were used. APES at room temperature for 15 h. The probe and target DNA were Figure 2a shows the thicknesses of the thin organic layers on then immobilized and hybridized on this aminated PDMS stamp silicon wafers when the silanization was performed in aqueous following the same procedure as that for the silicon wafers. solutions. It can be seen that the thicknesses of the organic layers After hybridization, the PDMS stamps with hybridized DNA on formed from APES and AEAPS increased with reaction time, their surfaces were brought into conformal contact with an APESwhich suggests that both APES and AEAPS formed multilayers coated glass slide for 30 s. The slide was wetted by DI water before on the silicon surfaces in aqueous solutions. In addition, after printing. A homemade printing device comprising two metal blocks was used to ensure that a constant pressure was applied to the entire 2 h of reaction, we observed that the water contact angle in surface. Figure 2b increased to 60°, which suggests that the surfaces Ellipsometry and Water Contact Angle Measurements. The either became more hydrophobic or became rougher after the film thickness on a surface was measured by using a Stoke formation of multilayers. Because multilayers formed from ellipsometer from Gaetner (Skokie, IL). The incident light was kept organosilanes are known to hydrolyze easily upon contact with at 70°, and a wavelength of 632.8 nm was used in all experiments. aqueous solutions, the immobilized DNA may detach along with Optical parameters n and k were determined by using a clean silicon wafer. For each sample, six different locations were measured. The the multilayers. Therefore, to prevent the formation of multilayers thickness was obtained by using software provided by the manuon the surface, we tested other conditions for the silanization facturer. process. The water contact angles of APES- and AEAPS-terminated silicon Since past studies18 have shown that methanol can be used as wafers were measured by using a VCA Optima goniometer from a solvent to inhibit the formation of multilayers on surfaces, a AST (Billerica, MA). A drop of water (1 µL) was placed on the surface of a silicon wafer, and the image of the droplet was captured mixture of methanol/water was used as a solvent to dissolve by using software provided by the manufacturer. APES and AEAPS for the silanization of the silicon wafers. Fluorescence Imaging and Quantification. The FAM-labeled Figure 2c shows the thicknesses of the organic layers formed DNA was observed with a fluorescence microscope (Eclipse E200, from APES and AEAPS as a function of the reaction time in Nikon, Tokyo, Japan). A fluorescein/enhanced green fluorescent methanol/water. It can be seen that the thickness of the APES protein (FITC/EGFP) filter set from Chroma Technology (Brattlefilm reached 6 Å after 6 h, which is consistent with the reported boro, VT) was used. A digital camera mounted on top of the microscope was used to capture images. The same parameters were thickness of an APES monolayer.18,19 For the AEAPS film, there used for all images and analyses. For quantitative fluorescence are two plateaus in the ranges of 3-6 and 15-21 h, corresponding measurements of FAM-labeled target DNA on the surfaces, the to thicknesses of 4 and 8 Å, respectively. Past studies have shown silicon wafers were immersed in 3 mL of KOH/KCl (130 mM/50 that the thickness of an AEAPS monolayer is either 4 Å20,21 or mM) for 7 min to elute the hybridized DNA. The fluorescence intensity of the solution was then measured by using a fluorescence spectrometer (Perkin-Elmer, Wellesley, MA). (18) Charles, P. T.; Vora, G. J.; Andreadis, J. D.; Fortney, A. J.; Meador, C.
Results and Discussion Surface Modification of Silicon Wafers with APES and AEAPS. First, we investigated the experimental conditions of forming amine-terminated surfaces for DNA immobilization.
E.; Dulcey, C. S.; Stenger, D. A. Langmuir 2003, 19, 1586-1591. (19) Moon, J. H.; Shin, J. W.; Kim, S. Y.; Park, J. W. Langmuir 1996, 12, 4621-4624. (20) Chrisey, L. A.; Lee, G. U.; O’Ferrall, C. E. Nucleic Acids Res. 1996, 24, 3031-3039. (21) Stenger, D. A.; Pike, C. J.; Hickman, J. J.; Cotman, C. W. Brain Res. 1993, 630, 136-147.
8610 Langmuir, Vol. 23, No. 16, 2007
5 ( 2 Å.22 In our case, a possible reason for the existence of two plateaus is that the secondary amine of AEAPS might adsorb on the surface in the beginning such that the thickness of the monolayer was only ∼4 Å; with the increase of the reaction time, the secondary amine desorbed from the surface and caused an increase in the thickness to 8 Å. This thickness is in reasonable agreement with a complete monolayer of fully extended AEAPS. Thus, we conclude that a complete monolayer of AEAPS can be formed on the silicon wafer after 12 h. Furthermore, we also conclude that monolayers of AEAPS and APES are easier to form on the silicon wafer surfaces by using a methanol/water mixture. The water contact angles of APES- and AEAPS-coated silicon wafers prepared in methanol/water are shown in Figure 2d. It can be seen that the contact angle decreased from approximately 50° to nearly 31° after 15-21 h of reaction time. Because the thickness did not change during the same period of time, the decrease in the water contact angle can be attributed to the reorientation of APES or AEAPS that gradually caused all the amine groups to point upward. It is also noted that the thickness of the thin film decreased and the water contact angle increased after 24 h of reaction time, which suggests that silanization is a very dynamic process. The hydrolyzed silane molecules not only adsorb/desorb on the silicon surfaces with hydrogen bonding and functionalize the surface, but they also react with each other in the solution phase to form oligomers. As a result, the concentration of free silane molecules in the solution decreases over time. The lower silane concentration will eventually lead to a decrease in the adsorption rate and cause the surface thickness to decrease. When the surface density of the immobilized silane becomes too low to form a complete monolayer, some molecules may lay down on the surface, leading to an increase in the surface roughness and water contact angle. Since a hydrophilic surface is preferred for the subsequent affinity microcontact printing of complementary DNA, all amineterminated surfaces were prepared in methanol/water for 15 h in the following experiments. Immobilization and Hybridization of DNA Molecules on the Aminated Silicon Wafer. Next, we study the influence of two different silanes (APES and AEAPS) on the immobilization of DNA and the subsequent DNA hybridization. In the first experiment, we immobilized FAM-labeled DNA probes on the APES- or AEAPS-modified surfaces and quantified the densities of DNA immobilized on these two surfaces. Parts a and b of Figure 3 show fluorescent images of APES- and AEAPS-modified silicon wafers after 2 µL of FAM-labeled DNA (20 µM D2) was immobilized on the surface. Comparing the fluorescence intensities of Figure 3a and b, it is found that the fluorescence intensity is stronger on the surface of the APES-modified silicon wafer than that on the surface of the AEAPS-modified silicon wafer. Thus, we conclude that a higher density of DNA can be achieved on the APES-modified surface. Next, to determine if a higher density of probe DNA can lead to a higher density of the target DNA after hybridization, we immobilized nonfluorescent DNA (D1) on the surface and then hybridized D1 with complementary FAM-labeled DNA (D2). The fluorescence intensities were measured by using a fluorescence microscope and a fluorescence spectrometer. In parts c and d of Figure 3, the fluorescence intensity from the APES-coated surface is stronger than that from the AEAPS-coated surface. Therefore, we conclude that APES is better than AEAPS as a silane coupling agent for DNA (22) Stenger, D. A.; Georger, J. H.; Dulcey, C. S.; Hickman, J. J.; Rudolph, A. S.; Nielsen, T. B.; McCort, S. M.; Calvet, J. M. J. Am. Chem. Soc. 1992, 114, 8435-8442.
Tan et al.
Figure 3. Fluorescence images of FAM-labeled DNA immobilized on (a) AEAPS- and (b) APES-modified silicon wafers. Fluorescence images of FAM-labeled target DNA hybridized with the probe DNA immobilized on (c) AEAPS- and (d) APES-modified silicon wafers.
immobilization and hybridization. Unfortunately, when noncomplementary FAM-labeled DNA (D3) was used in the DNA hybridization experiment, green fluorescence still could be seen on both surfaces, which suggests that, under the current immobilization and hybridization conditions, the hybridization of DNA is not specific. Thus, in the following experiments, we sought to find other hybridization conditions by measuring the fluorescence intensity of the hybridized target DNA on the surfaces. Figure 4a shows the fluorescence intensity of the hybridized DNA (D2 or D3) eluted from the surface. This result suggests that both complementary and noncomplementary target DNA can hybridize with the probe DNA, because the differences in fluorescence intensity are indiscernible. It also shows that DNA hybridization performed in SSC buffer does not provide much specificity for the complementary DNA. Thus, a new hybridization condition using TE buffer was attempted. Figure 4b shows that DNA hybridization performed in TE buffer led to a high specificity for complementary DNA for both 20-mer (D2 and D3) and 25-mer DNA (D5 and D6). A possible reason for the high specificity achieved by using TE buffer is the temperature; hybridization in SSC buffer was performed at 42 °C, whereas hybridization in TE buffer was performed at room temperature (25 °C). The dependence of specificity on the hybridization temperature can possibly be attributed to the length of the DNA. Because the DNA used in our study is relatively short (20- or 25-mers), we believe that a lower hybridization temperature is required to stabilize the double-helix structure. Although the hybridization protocol in SSC buffer is very common, it is probably more suitable for longer DNA. In addition, we do not think ionic strength is responsible for the high hybridization specificity in TE buffer, because the ionic strengths in 4× SSC and TE buffer are similar (0.65 M and 1 M, respectively). Finally, Figure 4b also shows that the fluorescence intensity ratio of complementary to noncomplementary for the 25-mer DNA is larger than that for the 20-mer DNA, which is consistent with past studies showing that the specificity of DNA hybridization increased with the number of base pairs of the DNA molecule. Affinity Microcontact Printing. After target DNA D2 hybridized with D1 immobilized on the surface of a PDMS stamp, we used ellipsometry to test if D2 can be transferred from the PDMS stamp to the surface of an aminated silicon wafer by
DNA Transfer Using Affinity Microcontact Printing
Langmuir, Vol. 23, No. 16, 2007 8611
Figure 4. Fluorescence intensity of hybridized target DNA (FAMlabeled) eluted from the surface of silicon wafers decorated with probe DNA. (a) Hybridization was carried out in SSC buffer with different strengths. (b) Hybridization was carried out in TE buffer, and two DNA strands with different numbers of base pairs were used.
using affinity microcontact printing. Since the patterned area on the PDMS stamp prepared from the silicon master was not large enough for ellipsometry measurements, we used a flat PDMS stamp for this experiment. In Table 1, the original thickness of the APES film on the surface of the silicon wafer was 6.2 Å. After the PDMS stamp with the complementary target DNA (D2) was printed on the APES-modified silicon wafer, the thickness of the printed area increased to 24.3 Å. To confirm that the increase was due to the transfer of DNA molecules, two control experiments were performed. First, we prepared one bare stamp and a stamp decorated with the probe DNA (D1) only and then printed them on an aminated silicon wafer. It was then found that the thicknesses on both surfaces increased to 13.5 and 15.1 Å, respectively. As a result, it is very difficult to confirm that the original increase of thickness (24.3 Å) is solely attributed to the transferred DNA molecules. We suspect that some unreacted monomers of PDMS may also transfer to the solid surface during contact printing, as suggested in a number of past studies.23-25 To overcome this problem, we employed a cleaning procedure26 to remove unreacted monomers from the PDMS stamp by extracting the PDMS stamp with ethanol. The results of the contact printing experiments using the “cleaned” PDMS stamps
are also shown in Table 1 for comparison. When the cleaned PDMS stamp with the hybridized target DNA was printed onto a silicon wafer, the thickness of the printed area increased from 6.1 to 16.1 Å. Two control experiments showed that the thicknesses after being printed by a bare PDMS stamp and a PDMS stamp with probe DNA remained at 6.2 and 6.3 Å, respectively. It suggests that the transfer of unreacted monomers was successfully prevented by extracting the PDMS in ethanol prior to use. Thus, we conclude that the increase from 6.1 to 16.1 Å after the surface was printed by using a PDMS stamp with hybridized DNA can be attributed to the transfer of DNA molecules. Finally, fluorescence microscopy was used to study the transfer of DNA to a solid surface by using affinity microcontact printing. We prepared two PDMS stamps with immobilized probe DNA (D1) on their surfaces. Complementary (D2) and noncomplementary (D3) DNA strands were then hybridized with D1 on the surfaces of the PDMS stamps. After microcontact printing of the PDMS stamps with D2 and D3 on glass slides, the fluorescent images on the glass slides were recorded. Figure 5a shows that no fluorescence was detected on the surface printed with the PDMS stamp hybridized with noncomplementary DNA. In contrast, patterned fluorescence images can be observed on the surface printed with the PDMS stamp hybridized with the complementary DNA (Figure 5b). This suggests that noncomplementary target DNA was not transferred from the PDMS stamp to the glass surface. Parts c and d of Figure 5 show the fluorescence intensity profiles of parts a and b, respectively, along the dotted line. It can be seen that the fluorescence intensity in Figure 5c is very flat and low, suggesting that no FAMlabeled DNA was present on the surface. In contrast, two plateaus in Figure 5d arise from the regions with transferred FAM-labeled DNA molecules on the surface. It is also noted that the two plateaus are relatively smooth, which implies that DNA transfer during the printing process was quite uniform. To eliminate the possibility that the entire DNA duplex is transferred from the PDMS stamp to the amine-terminated glass slide upon contact printing, we performed the following experiments. First, fluorescence-labeled probe DNA (D5) was immobilized on the amine-terminated PDMS stamp and then hybridized with a nonfluorescence-labeled complementary DNA (D4). As shown in Figure 6a, the fluorescent image of the PDMS stamp indicates the presence of the fluorescence-labeled probe DNA on the surface. However, after the PDMS stamp was printed onto an amine-terminated glass slide, no green fluorescence could be detected on the surface of the glass slide (Figure 6b), whereas the fluorescence intensity on the PDMS stamp remained unchanged (Figure 6c). This result implies that the fluorescencelabeled probe DNA adhered to the stamp surface tightly and was not transferred to the amine-terminated glass slide during the contact printing procedure. It is also consistent with our ellipsometry data (Table 1) showing that the surface thickness of the aminated glass slide did not increase after the slide was printed with a PDMS stamp with the immobilized DNA probe alone. Next, when we replaced the fluorescence-labeled probe
Table 1. Thicknesses (Å) of Surface Organic Layers on Silicon Wafers before and after Being Printed with Different PDMS Stamps
PDMS stamps as prepared PDMS stamps extracted with ethanol
before printing
after printing with a PDMS stamp decorated with hybridized DNA
after printing with a bare PDMS stamp
after printing with a PDMS stamp decorated with probe DNA
6.2 ( 0.1
24.3 ( 6.8
13.5 ( 5.1
15.1 ( 5.9
6.1 ( 0.1
16.1 ( 4.7
6.2 ( 0.7
6.3 ( 1.1
8612 Langmuir, Vol. 23, No. 16, 2007
Tan et al.
Figure 5. Fluorescence images of aminated silicon wafers printed with PDMS stamps decorated with target DNA (a) when the noncomplementary target DNA was used and (b) when the complementary target DNA was used. Before printing, the stamp was incubated in TE buffer with FAM-labeled target DNA, which hybridized with the immobilized probe DNA on the surface. (c) and (d) are the fluorescence intensity profiles from (a) and (b), respectively, along the dashed line.
Figure 6. Fluorescence images of DNA on the PDMS stamps (a, c, d, and f) and on the amine-terminated glass slides (b and e). These images were used to monitor the transfer of DNA from the PDMS stamps to the amine-terminated glass slides. (a) An amine-terminated PDMS stamp with immobilized fluorescence-labeled probe DNA hybridized with nonfluorescence complementary DNA. (b) An amine-terminated glass slide onto which the PDMS stamp in (a) was printed. (c) The PDMS stamp in (a) after printing on the aminated glass slide. (d) An amine-terminated PDMS stamp with immobilized nonfluorescence-labeled probe DNA hybridized with fluorescence-labeled complementary DNA. (e) An amine-terminated glass slide onto which the PDMS stamp in (d) was printed. (f) The PDMS stamp in (d) after printing on the aminated glass slide.
DNA (D5) with nonfluorescence-labeled probe DNA (D4) and hybridized it with fluorescence-labeled target DNA (D5), green fluorescence could be detected on the stamp surface (Figure 6d). After the stamp was printed onto an amine-terminated slide, we observed that regions of green squares were successfully transferred from the PDMS stamp (Figure 6f) to the glass slide as shown in Figure 6e. These results, when combined, lead us to conclude that only the hybridized DNA target, but not the
immobilized DNA probe, was transferred to the aminated glass slide during the contact printing procedure. We propose that the transfer of DNA is based on the following mechanism. When a PDMS stamp decorated with double-stranded DNA is in contact with the surface of an amine-terminated silicon wafer, the electrostatic attraction between the negative DNA molecules and positive amine groups is stronger than the hydrogen
DNA Transfer Using Affinity Microcontact Printing
Langmuir, Vol. 23, No. 16, 2007 8613
off the surface, these bonds will break one by one, depending on their strength. Otherwise, the stamp will not be separated from the glass slide. The first one to break is probably the hydrogen bond (2) because it has the lowest binding energy. Next, the weakest ionic bonds, 4R for example, will break as shown in Figure 7b. The procedure will continue until a few ionic bonds are left as shown in Figure 7c. In this case, the ionic bonds (4β) are strong enough to pull the target DNA away from the probe DNA immobilized on the other surface. As a result, the target DNA is transferred onto the aminated glass slide as shown in Figure 7d. We can also consider a more complicated scenario where the probe DNA is immobilized on the surface with two covalent bonds (or strong electrostatic interactions) as shown in Figure 7e. The target DNA also hybridizes with the probe DNA. Similarly, when the stamp is removed from the surface, the hydrogen bond (2) may break first, leading to Figure 7f. In this case, the ionic bonds (4β) cannot pull the probe DNA away because of the entanglement in the central region. To remove the probe DNA from the stamp surface, only one anchor on the surface is allowed. Therefore, one of the two 4β ionic bonds needs to be broken before the target DNA can be pulled away. For instance, if one of the 4β ionic bonds breaks, then the other 4β ionic bond can be responsible for transferring the DNA to the aminated surface as shown in Figure 7g and h. For noncomplementary DNA, since only a limited amount of noncomplementary DNA can hybridize with the probe DNA molecules, the amount of transferred target DNA on the surfaces was very low. Figure 7. Proposed model for the transfer of complementary DNA from an aqueous solution onto an aminated surface by using affinity microcontact printing. (a-d) show the first scenario where only one end of the DNA is immobilized on the PDMS stamp surface and (e-h) show the second scenario where both ends of the probe DNA are immobilized on the PDMS stamp surface. (1) represents a covalent bond (or strong electrostatic interactions), and (2) represents a hydrogen bond holding the double helix. (3) and (4) represent the electrostatic attraction between the negatively charged DNA and the positively charged surface. Letters R and β represent the surfaces with which (PDMS stamp and glass slide, respectively) the DNA interacts. The black and gray color strands are probe DNA and target DNA, respectively.
bonds holding the double-helix structure of DNA. As a result, the hybridized target DNA molecules are transferred from the PDMS stamps to the silicon wafer surfaces. The DNA transfer mechanism can be better illustrated in Figure 7. Figure 7a shows immobilized DNA hybridized with complimentary target DNA on the surface of a PDMS stamp. There are many bonds formed between the two strands of DNA and between the DNA and two aminated surfaces, such as the covalent bonds (or strong electrostatic interactions) between the probe DNA and PDMS stamp (1), the hydrogen bonds holding the double helix (2), and several ionic bonds (3 and 4) between the negatively charged DNA and positively charged aminated surfaces. Under this condition, the double helix cannot be separated without untangling and unwinding the helix. Nevertheless, when the stamp is lifting (23) Kumar, A.; Biebuyck, H. A.; Whitesides, G. M. Langmuir 1994, 10, 1498-1511. (24) Graham, D. J.; Price, D. D.; Ratner, B. D. Langmuir 2002, 18, 15181527. (25) Glasmastar, K.; Gold, J.; Andersson, A. S.; Sutherland, D. S.; Kasemo, B. Langmuir 2003, 19, 5475-5483. (26) Raiber, K.; Terfort, A.; Benndorf, C.; Krings, N.; Strehblow, H. H. Surf. Sci. 2005, 595, 56-63.
Conclusion In this paper, the transfer of DNA molecules from PDMS stamps to amine-terminated surfaces by using affinity microcontact printing has been successfully demonstrated. We propose that the transfer mechanism is based on the interaction between the positively charged amine groups and the negatively charged DNA that is stronger than the hydrogen bonds between two hybridized DNA molecules. Since amine-terminated monolayers on the surfaces are desired for the subsequent affinity microcontact printing, we compare two silane coupling agents, AEAPS and APES, for the chemical modification of silicon wafers and study the effects of different solvents. It is concluded that the surfaces modified with APES in methanol/water are more suitable for the immobilization of probe DNA on the surfaces. We also compare two hybridization conditions: one in SSC buffer at 42 °C and the other in TE buffer at room temperature. We show that DNA hybridization performed in SSC buffer does not lead to any specificity for complementary DNA, whereas DNA hybridization performed in TE buffer yields high specificity for complementary DNA. Finally, we show that the transfer of unreacted monomers from the PDMS stamp can be prevented by extracting the stamp in ethanol prior to use. The study reported here provides a fast and simple method for the transfer of specific DNA molecules from aqueous solution onto a solid surface. It is believed that this technique has the potential to be used as a means of detecting DNA molecules with good specificity. Acknowledgment. This work is funded by the Agency for Science and Technology Research (A*STAR) in Singapore under Project No. 0521010099. LA701258C
