Systematic Investigation of Optimal Aptamer Immobilization for Protein

Aug 27, 2008 - Tobias Fischer , Paul M. Dietrich , Cornelia Streeck , Santanu Ray ... Alex Shard , Burkhard Beckhoff , Wolfgang E. S. Unger , and Knut...
2 downloads 0 Views 848KB Size
Download PDF
Anal. Chem. 2008, 80, 7372–7378

Systematic Investigation of Optimal Aptamer Immobilization for Protein-Microarray Applications ¨ znur Ko¨kpinar, Karl Friehs,† Frank Stahl,* and Johanna-Gabriela Walter, O Thomas Scheper Institut fu¨r Technische Chemie, Leibniz Universita¨t Hannover, Callinstrasse 3, 30167 Hannover, Germany Aptamers are short single-stranded DNA or RNA oligonucleotides that can bind to a wide range of target molecules with high affinity and specificity. As nucleic acids, aptamers can undergo denaturation, but the process is reversible. As a result of this stability and the possibility of automated selection of aptamers, these oligonucleotides are highly promising capture molecules in microarray formats. In this study, his-tagged proteins and an aptamer directed against the his-tag were chosen as a model system. Different factors affect the activity of aptamers immobilized on a solid support like a microarray surface. The orientation of the immobilized aptamer plays an important role in correct aptamer folding and, thus, in effective binding of the corresponding target. Other important parameters identified in this work are the microarrays’ surface charge as well as the length of the spacer between aptamer and solid support. These parameters were investigated systematically, resulting in the development of an aptamer-based microarray for detection of his-tagged proteins. The general applicability of the developed immobilization strategy was demonstrated by utilization of three different aptamers. Protein microarrays are promising tools for postgenomic research. Today, most protein microarray approaches are based on the recognition of the target molecule by immobilized antibodies.1-3 One disadvantage of antibodies in protein microarray technology is their lack of stability since antibodies tend to denature when immobilized on a solid support. As a consequence, the routine use of antibody arrays is limited. As nucleic acids, aptamers can undergo denaturation as well, but the process is reversible.4 Aptamers are short single-stranded DNA or RNA oligonucleotides that can bind to a wide range of target molecules with high affinity and specificity. Like antibodies, aptamers are able to fold into a well-defined three-dimensional structure that allows the binding of the corresponding protein, * To whom correspondence should be addressed. E-mail: stahl@ ift.uni-hannover.de. † Current address: AG Fermentationstechnik, Universita¨t Bielefeld, Technische Fakulta¨t, 33594 Bielefeld, Germany. (1) Kusnezow, W.; Hoheisel, J. D. Biotechniques 2002, 14–23. (2) Glokler, J.; Angenendt, P. J. Chromatogr., B: Anal. Technol. Biomed. Life Sci. 2003, 797 (1-2), 229–40. (3) Angenendt, P. Drug Discovery Today 2005, 10 (7), 503–11. (4) Jayasena, S. D. Clin. Chem. 1999, 45 (9), 1628–50.

7372

Analytical Chemistry, Vol. 80, No. 19, October 1, 2008

resulting in similar or even higher affinities than those of antibodies.5,6 Therefore, aptamers are beginning to rival antibodies in many applications based on molecular recognition such as ELISA, biosensors, and protein microarrays.4,5,7-11 Another advantage of aptamers over antibodies is the broad range of possible target molecules. While the antibody identification process depends on animals or cell culture, aptamers are selected by an in vitro selection and amplification technique known as SELEX (systematic evolution of ligands by exponential enrichment),12,13 which allows the development of aptamers directed against virtually all kinds of target molecules including toxins14 or less immunogenic substances. Moreover, the functionality of aptamers is not dependent on physiological conditions which is a limiting factor in antibody microarray procedures. In the case of aptamers, the conditions during SELEX, including buffer composition, pH, and temperature, can be selected in regard to intended assay conditions. Consequently, it is not necessary to adapt the assay to the molecular recognition element. Furthermore, the aptamer itself can be customized in order to fit into desired applications. Following the SELEX procedure, aptamers can be produced chemically with high reproducibility and minimal batchto-batch variation. Furthermore, it is possible to modify the aptamers at accurately defined positions. Due to this fact and the possibility of automated selection of aptamers, these oligonucleotides are highly promising capture molecules in immunoassay formats. Compared to antibody microarrays, aptamer microarrays are at a early stage. Until now, only several attempts have been made to develop aptamer-based protein microarrays. In most publicized works, the aptamers are bound via the steptavidin-biotin interaction.10,15,16 Due to the limited stability of streptavidin, this (5) Hesselberth, J.; Robertson, M. P.; Jhaveri, S.; Ellington, A. D. J. Biotechnol. 2000, 74 (1), 15–25. (6) Brody, E. N.; Gold, L. J. Biotechnol. 2000, 74 (1), 5–13. (7) Baldrich, E.; Restrepo, A.; O’Sullivan, C. K. Anal. Chem. 2004, 76 (23), 7053. (8) O’Sullivan, C. K. Anal. Bioanal. Chem. 2002, 372 (1), 44–8. (9) Bini, A.; Minunni, M.; Tombelli, S.; Centi, S.; Mascini, M. Anal. Chem. 2007, 79 (7), 3016–9. (10) Collett, J. R.; Cho, E. J.; Ellington, A. D. Methods 2005, 37 (1), 4–15. (11) You, K.; Lee, S.; Im, A.; Lee, S. Biotechnol. Bioprocess Eng. 2003, 8 (2), 64–75. (12) Tuerk, C.; Gold, L. Science 1990, 249 (4968), 505–10. (13) Ellington, A. D.; Szostak, J. W. Nature 1990, 346 (6287), 818–22. (14) Bruno, J. G.; Kiel, J. L. Biotechniques 2002, 32 (1), 178–80. (15) Cho, E.; Collett, J.; Szafranska, A.; Ellington, A. Anal. Chim. Acta 2006, 564 (1), 82–90. 10.1021/ac801081v CCC: $40.75  2008 American Chemical Society Published on Web 08/27/2008

approach does not exploit the advantages of aptamer stability resulting in a shorter lifetime of spotted microarrays. Further investigations in regard to optimization of immobilization chemistry and assay conditions have to be done. In this study, aptamers were immobilized on amino-modified surfaces in a covalent manner without using streptavidin. This should result in a more stable microarray that does not require chilled storage. The influence of aptamer orientation, spacer length, and surface charge on aptamer microarray performance was investigated systematically. His-tagged proteins and an aptamer directed against the his-tag were chosen as a model system. The general applicability of the developed immobilization strategy was demonstrated by utilization of three different aptamers. EXPERIMENTAL SECTION Materials. Anti-his-tag aptamers (6H7, 5′-GCT ATG GGT GGT CTG GTT GGG ATT GGC CCC GGG AGC TGG C-3′, and 6H5, 5′-GGC TTC AGG TTG GTC TGG TTG GGT TTG GCT CCT GTG TAC G-3′) and anti-streptavidin aptamer (miniStrep, 5′-TCT GTG AGA CGA CGC ACC GGT CGC AGG TTT TGT CTC ACA G-′3)17 as well as a defective sequence of miniStrep (miniStrepdef, 5′-TCT GTG AGA CGA CGC ACC GTC GCA GGT TTT GTC TCA CAG′3) were obtained from Operon and were 5′- and optionally 3′modified with amino groups during synthesis. The 6H7 and the 6H5 aptamer sequences were taken from the US patent specification U.S. 2005/0142582 A1.18 The aptamers are directed against the his-tag and were selected in 50 mM K2HPO4, 150 mM NaCl, 0.05% Tween 20, pH 7.5 (PBST-6H7, without Tween 20: PBS-6H7). The anti-streptavidin aptamer was selected by Bittker et al. via nonhomologous random recombination utilizing 150 mM NaCl, 10 mM MgCl2, 50 mM Tris-Cl, pH 8.0 (miniStrepBP) as selection buffer.17 The defective sequence miniStrepdef was attained by skipping one base of the miniStrep aptamer. Protein Expression and Purification. The Escherichia coli strain JM109 pET-bgl-his for production of His6-tagged β-glucanase (Bgl-His, MW 26.700, pI 6.1) was kindly donated by Prof. Dr. K. Friehs, AG Fermentationstechnik, University of Bielefeld. The cultivation was performed in shaking flasks with 100 mL of LB medium (10 g/L casein peptone, 10 g/L NaCl, 5 g/L yeast extract) supplemented with 100 µg/mL ampicillin. At an optical density, OD600, of 0.8-1.0, the expression of Bgl-His was induced by addition of isopropyl β-D-thiogalactopyranoside to a final concentration of 1 mM. The culture was grown overnight at 30 °C. The cells were harvested by centrifugation at 3345g for 30 min (4 °C). As an extracellular protein, Bgl-His was purified from fermentation broth via immobilized metal chelate chromatography utilizing iminodiaccetic acid (IDA)-modified membrane adsorber devices. A screening for the optimal metal ion for purification of Bgl-His was performed as described before,19 resulting in the best purification utilizing Co2+ immobilized on IDA membrane. The purification of Bgl-His was performed utilizing Sartobind IDA 75 (16) Collett, J. R.; Cho, E. J.; Lee, J. F.; Levy, M.; Hood, A. J.; Wan, C.; Ellington, A. D. Anal. Biochem. 2005, 338 (1), 113–23. (17) Bittker, J. A.; Le, B. V.; Liu, D. R. Nat. Biotechnol. 2002, 20 (10), 1024–9. (18) Doyle, M. B.; Murphy, S. A. Patent Application Publication, US 2005/ 0142582 A1, 2005. (19) Harkensee, D.; Kokpinar, O.; Walter, J.; Kasper, C.; Beutel, S.; Reif, O. W.; Scheper, T.; Ulber, R. Eng. Life Sci. 2007, 7 (4), 388–94.

units (Sartorius Stedim Biotech, Go¨ttingen, Germany) as recommended by the manufacturer. In order to remove imidazole from the purified protein, a buffer exchange of the eluate was performed (3 kDa MWCO, Vivaspin 20, Satorius Stedim Biotech, Go¨ttingen, Germany). Protein Labeling. Purified proteins were labeled with Cy3 Mono-Reactive Dye (Amersham Biosciences, Piscataway, NJ). In contrast to the recommendations of the manufacturer, one vial of the dye was used to label 0.2 mg of protein. Unreacted dye was removed by excessive ultrafiltration (3 kDa MWCO, Microcon 3, Millipore GmbH, Schwalbach, Germany) and subsequent desalting with Protein Desalting Spin Colums (Pierce Biotechnology, Rockford, IL). The protein and dye concentrations were determined with the Nanodrop ND-1000 spectrophotometer (NanoDrop Technologies, Inc., Wilmington, DE). The concentrations were used to calculate the dye to protein (D/P) ratio, resulting in a D/P of 0.75 for Cy3-labeled Bgl-His. Cy3-labeled streptavidin was purchased from Jackson Immuno Research Laboratories (Suffolk, UK). Aptamer Activation. Aptamers were activated with cyanuric chloride according to the protocol for activation of oligonucleotides published by Ferguson et al.20 In brief, the aptamer (50 nmol) was diluted in 450 µL of SBB (0.1 M sodium borate buffer, pH 8.3) and 25 nmol of cyanuric chloride was added in 50 µL of acetonitrile. After 1-h reaction at room temperature, unreacted cyanuric chloride was removed by ultrafiltration (3 kDa MWCO, Microcon 3, Millipore). Furthermore, an optimization of this protocol was performed by variation of the molar ratio of aptamer and cyanuric chloride (1:0.5; 1:1; 1:5; 1:100). For the aptamers directed against the his-tag (6H5 and 6H7), the activation was performed in SBB as described above and additionally in the buffer system the aptamers were selected in (50 mM K2HPO4, 150 mM NaCl, pH 7.5). After spotting, the immobilization efficiency was estimated utilizing Sybr Green II nucleic acid stain and the functionality of the aptamer by incubation with the fluorescently labeled target protein. Preparation of PEI-Modified Glass Slides. Aldehydemodified glass slides (VSS-25 Silylated Slides, CEL Associates, Inc., Pearland, TX) were incubated in 5% polyethylenimine (PEI) at room temperature for 1 h. Consequently, the slides were washed with PBS for 5 min twice and dried with compressed air. Prior to spotting, the slides were stored overnight in the dark. In addition to the custom-made slides, commercially available aminomodified microarray substrates (UltraGAPS Coated Slides, Corning, Acton, MA) were used. Aptamer Immobilization. Aptamers were spotted on PEImodified glass slides or on UltraGAPS Coated Slides in the same buffer they were activated in (SBB or PBS-6H7, respectively). Dilution rows of the activated aptamers were spotted in 10 replicates of every dilution using a 427 arrayer (Affymetrix, Santa Clara, CA) equipped with 125-µm solid pins. Spotting was performed by placing five hits per dot. After spotting, the slides were incubated at room temperature overnight in the dark. Capping of Amino Modifications. The printed microarrays were incubated in a fresh solution of 0.275 g of succinic acid in (20) Ferguson, J. A.; Boles, T. C.; Adams, C. P.; Walt, D. R. Nat. Biotechnol. 1996, 14 (13), 1681–4.

Analytical Chemistry, Vol. 80, No. 19, October 1, 2008

7373

Scheme 1. (A) Activation of Aptamer Utilizing Cyanuric Chloride. (B) Coupling of Activated Aptamer to Amino-Modified Glass Slides. (C) Capping of Remaining Amino Groups with Succinic Anhydride

17 mL of DMSO and 1 mL of 0.1 M NaHCO3, pH 9.4, for 5 min. Consequently, the slides were washed with ddH2O for 1 min twice. Determination of Immobilization Efficiency. The density of the immobilized aptamer was estimated with Sybr Green II stain specific for single-stranded nucleic acids. Prior to the stain, the aptamers were denaturized by placing the slide into boiling ddH2O. In order to avoid a refolding of the denaturized aptamer, the slide was instantly dried by compressed air. The slide was incubated in a 1:10 000 solution of Sybr Green II in TAE (40 mM Tris, 1 mM EDTA, 20 mM acetic acid, pH 8.5). Afterward the slide was washed with TAE and ddH2O for 1 min, respectively. Prior to scanning, the slide was dried with compressed air. Binding Assay. After blocking for 30 min (1% BSA in selection buffer) and washing in selection buffer (5 min), the aptamers were denaturized by incubation in boiling ddH2O (1 min). Afterward, the slides were incubated in selection buffer for 30 min in order to obtain the correct aptamer folding. For target binding, the slides were incubated with fluorescently labeled protein (800 µL, 5 µg/mL Cy3-labeled Bgl-His in PBST6H7 or 5 µg/mL Cy3-labeled streptavidin in miniStrepBP) for 4 h at 20 °C utilizing a Secure Seal Hybridization Chamber (Grace Biolabs, Bend, OR). During the incubation,the slides were shaken at 300 rpm (Thermomixer comfort supplemented with slide adapter, Eppendorf, Hamburg, Germany). Prior to scanning, the slides were washed with selection buffer for 5 min twice and dried with compressed air. Scanning and Analysis. Scanning was performed using a GenePix4000B (Axon Instruments, Foster City, CA). ImaGene 5 (BioDiscovery, El Segundo, CA) software was used to analyze the images. Mean values of all replicates were used for calculations. Relative signal intensity was calculated as the difference of signal 7374

Analytical Chemistry, Vol. 80, No. 19, October 1, 2008

mean (SM) and background mean (BM). Signal-to-noise ratio (SNR) was calculated by

SNR )

(SM - BM) standard deviation of background

Limit of detection was defined as SNR ) 3.21 In this study, optimal PMT settings were determined by variation of gain settings in intervals of 100. For data analysis, PMT settings were chosen that result in highest SNR values. RESULTS AND DISCUSSION We have systematically investigated and optimized different immobilization parameters in regard to aptamer activity. Utilizing high-reactive aldehyde- or epoxy-modified surfaces for immobilization of aptamers may not only result in attachment of oligonucleotides via the terminal amino linker but also in attachment via nucleobases within the aptamer sequence. In order to ensure a more controlled immobilization via the terminal NH2 linker, the aptamers were activated with cyanuric chloride and immobilized on rather unreactive amino-modified surfaces (Scheme 1A and B). According to Van Ness et al., the activation with cyanuric chloride exclusively occurs at the terminal amino-linker modification of oligonucleotides, thus enabling a highly directed immobilization.22 The influence of the microarrays’ surface charge, the aptamers’ orientation, and the spacer length were investigated. Furthermore, the activation procedure was optimized. The method was developed utilizing the 6H7 aptamer directed against the his(21) Verdnick, D.; Handran, S.; Pickett, S., Key considerations for accurate microarray scanning and image analysis. In DNA image analysis: nuts and bolts; Kamberova, G., Ed.; DNA Press: Salem, MA, 2002. (22) Van Ness, J.; Kalbfleisch, S.; Petrie, C. R.; Reed, M. W.; Tabone, J. C.; Vermeulen, N. M. Nucleic Acids Res. 1991, 19 (12), 3345–50.

Figure 1. Effect of surface charge. (A) 6H7 aptamer immobilized on UltraGAPS slides without subsequent capping of amino functions. (B) 6H7 aptamer immobilized on UltraGAPS slides with capping of amino functions. Numbers indicate spotting scheme: (1) 3′-modified aptamer, 100 µM; (2) 3′-modified aptamer, 10 µM; (3) 3′-modified aptamer, 1 µM; (4) 5′-modified aptamer, 100 µM; (5) 5′-modified aptamer, 10 µM; (6) 5′-modified aptamer, 1 µM.

tag. General applicability of the developed immobilization procedure was demonstrated by transferring the method to two different aptamers. Influence of Surface Charge on Analyte Binding Assay. Electrostatic interactions between aptamer and surface may interfere with correct aptamer folding and thus with binding of analyte. In order to investigate the influence of the surface charge on aptamer activity, 6H7 aptamers were spotted on UltraGAPS-modified slides. One slide was incubated with Cy3labeled Bgl-His without capping of amino groups (Figure 1A), one slide after subsequent capping of amino groups (Figure 1B). On UltraGAPS without capping, the 5′-modified 6H7 shows activity but the signal intensity is rather low. Since aminomodified surfaces exhibit a positive net charge, the aptamers’ negative phosphate backbone can interact with the surface electrostatically, resulting in an unfolding of aptamer on the surface and thus in a loss of activity. The capping of amino groups with succinic anhydride converts the positive amino groups into negative carboxy groups. The resulting negative charge of the surface prevents unfolding of aptamers due to electrostatic repulsion of the negative nucleic acids. In accordance with this considerations, the signal intensity increases after capping with succinic anhydride (Figure 1). Furthermore, the background decreases after capping resulting in higher SNR values. The reduced background fluorescence can be explained by more effective blocking of the distinct negative surface compared to the less positive GAPS surface. On GAPS-modified surfaces, the activity of 6H7 is highly dependent on the aptamer orientation (Figure 1). While the 5′-modified aptamer shows target binding, the 3′-modified aptamers shows no activity. This orientation seems to interfere with correct aptamer folding due to steric hindrance. GAPSmodified surfaces only provide a relative short spacer. Due to steric hindrance, the aptamers immobilized in proximity to the surface may not be able to fold into the correct conformation. By utilization of a longer spacer, the steric hindrance should

be minimized. Thus, experiments with PEI-modified surfaces were performed. Influence of Aptamer Orientation and Spacer Length. Due to steric hindrance, aptamers immobilized close to the surface may not be able to fold into the three-dimensional structure necessary for target recognition. The utilization of a spacer between aptamer and surface can reduce this effect. To investigate the influence of the spacer length, comparative experiments were performed with 6H7 spotted on GAPSmodified surfaces providing a short spacer and PEI-modified surfaces providing a huge spacer. Prior to binding studies, capping of amino groups was performed. On UltraGAPS slides, a high signal intensity along with a low background was observed, resulting in excellent SNR of 180 (Figure 2A and C). Utilizing PEI-modified surfaces, the signal intensity was even higher (Figure 2B). PEI not only provides a huge spacer but also increases the binding capacity of the surface.23,24 As a consequence of higher background, the SNR is lower than on UltraGAPS (Figure 2C). As described above, the 3′-modified aptamer shows no activity on UltraGAPS slides. Consequently, the functionality of the 6H7 aptamer depends on the aptamer orientation when a short spacer is used. In contrast, the activity of the aptamer shows no dependency on the immobilization orientation on PEI slides. The additional huge spacer ensures correct folding of the 6H7 aptamer. Thus, PEI-modified surfaces are suitable for immobilization of aptamers with a higher steric demand. Optimization of Aptamer Immobilization. The activation of aptamers was optimized by variation of the molar ratio of aptamer and cyanuric chloride (1:0.5; 1:1; 1:5; 1:100). Activation and immobilization were performed in SBB. The aptamers were spotted onto PEI-modified slides in concentrations from 1 to 250 µM, and the microarrays were processed as described above. In addition to target binding studies, the immobilization efficiency was investigated with Sybr Green II. The signal intensity of Sybr Green II stained slides shows high dependency on molar ratio of aptamer to cyanuric chloride (Figure 3A). High concentrations of cyanuric chloride during the activation result in high signal intensities, indicating high densities of aptamers on the surface. The activation efficiency and thus the immobilization efficiency are best at high concentrations of cyanuric chloride. To investigate the influence of different activation procedures on aptamer activity, target binding studies were performed (Figure 3B). At low spotted concentrations (e100 µM), the aptamer activated with 100-fold excess of cyanuric chloride shows good signal intensities. At higher concentrations, the signal intensities decrease. Immobilization of high densities of aptamers seems to interfere with correct aptamer folding due to steric hindrance or intermolecular base pairing. While the density of aptamer is high when utilizing 100-fold excess of cyanuric chloride, for target binding, a 5-fold excess performs better with highest relative signal intensities and a nearly linear correlation between aptamer concentration and signal intensities (Figure 3B). (23) Spiridonova, V. A.; Kopylov, A. M. Biochemistry (Moscow) 2002, 67 (6), 706–9. (24) Lee, M.; Walt, D. R. Anal. Biochem. 2000, 282 (1), 142–6.

Analytical Chemistry, Vol. 80, No. 19, October 1, 2008

7375

Figure 2. Influence of aptamer orientation and spacer length. (A) 6H7 aptamer immobilized on UltraGAPS surface. (B) 6H7 aptamers immobilized on PEI-modified surface. (C) SNR analysis. Spotting scheme is equivalent to Figure 1.

Figure 3. Optimization of immobilization procedure. Influence of molar aptamer to cyanuric chloride ratio on immobilization efficiency (A) and target binding (B).

Typically, the activation of amino-modified oligonucleotides with cyanuric chloride is performed in SBB.22 We have tested the suitability of the aptamer selection buffer (PBS-6H7) for activation procedure and compared it to the standard activation procedure in SBB. The 6H7 aptamer was activated with cyanuric acid in the optimal molar ratio of aptamer to cyanuric 7376

Analytical Chemistry, Vol. 80, No. 19, October 1, 2008

Figure 4. Optimization of immobilization by variation of activation buffer. (A) Influence of buffer on immobilization efficiency. (B) Influence of buffer on binding assay.

chloride (1:5) in SBB and in PBS-6H7. The influence of the buffer on immobilization efficiency is shown in Figure 4A. As expected, the utilization of basic SBB results in higher immobilization efficiencies. In contrast, the aptamers activated and immobilized in PBS-6H7 show higher signals in target binding assays (Figure 4B). While the amount of immobilized aptamer may be low under these conditions, the density of

Figure 5. Application of developed immobilization strategy to other aptamers. (A) 6H7. (B) 6H5. 6H7 and 6H5 are selected in the same SELEX procedure. (C) Ministrep aptamer. This aptamer was selected under different SELEX conditions.

functional aptamer seems to be higher. The aptamers activated in PBS-6H7 show higher signal intensities in target binding assays, indicating that the aptamers are immobilized in a more functional conformation when activated in their selection buffer. These results suggest that the aptamer may fold into its correct three-dimensional structure in PBS-6H7 during the activation procedure and is able to retain this structure after immobilization. To ensure this assumption, further experiments were performed. The aptamers activated in the different buffer systems (SBB and PBS-6H7) were used in target binding studies with and without denaturation and folding of the immobilized aptamers prior to incubation with Cy3-labeled BglHis. The results are shown in Figure 4B. In accordance with the assumption stated above, the aptamer activated in PBS6H7 shows good target binding without previous denaturation

and folding of the aptamer. In contrast, the aptamers activated in SBB show no activity without previous denaturation and folding that ensures correct aptamer conformation. These results confirm that the aptamers are totally or partially folded into the correct three-dimensional structure during activation in their selection buffer PBS-6H7. Application of the Developed Immobilization Strategy to Other Aptamers. In order to investigate the general applicability of the developed immobilization strategy, experiments were performed utilizing different aptamers. The method was developed using 6H7, which was selected against the his-tag. The method was first applied to 6H5; another aptamer directed against the his-tag emanated from the same selection as the 6H7 aptamer.18 Furthermore, experiments were performed with an aptamer directed against streptavidin (referred to as miniStrep), which was selected under different conditions.17 6H5 and 6H7 aptamers were activated in PBS-6H7 and were spotted on the same microarray slide. The microarray was processed as described above. While the 6H7 aptamer shows higher signal intensities, both aptamers show functionality on the same slide (Figure 5A and B). The activation and immobilization procedure was directly transferable to the 6H5 aptamer without any modifications. The transfer of the method to the miniStrep aptamer requires slight modification of the developed method. Since the aptamer was selected in a Tris-based buffer system (150 mM NaCl, 10 mM MgCl2, 50 mM Tris-Cl, pH 8.0), the activation with cyanuric acid cannot be carried out in the selection buffer. Consequently, the activation and immobilization of miniStrep was performed in SBB. Furthermore, the processing of the microarrays had to be adopted to the conditions utilized during miniStrep selection. The immobilized aptamer was denatured in boiling water (5 min), and the slides were subsequently transferred to ice-water for 5 min before the correct aptamer folding was achieved by incubation in the selection buffer. After these modifications, the miniStrep aptamer shows positive target binding (Figure 5C). In order to ensure that the binding between miniStrep and Cy3-labeld streptavidin depends on the special three-dimensional structure of the aptamer rather than on unspecific binding, comparative experiments were performed with miniStrep and miniStrepdef immobilized on PEI slides. Although the miniStrep aptamer and the miniStrepdef sequence only differ in one base, the miniStrepdef shows no target binding (data not shown). CONCLUSIONS We have developed and optimized a method that enables the covalent attachment of aptamers to microarray surfaces without affecting correct aptamer folding. Due to utilization of unreactive amino-modified surfaces and cyanuric chloride-activated aptamers, a controlled immobilization via the terminal amino linker of the oligonucleotide was achieved. In this study, the following factors concerning aptamer activity were investigated: (1) The orientation of the aptamer plays an important role in correct aptamer folding and thus in effective target binding. (2) Utilization of a huge spacer between immobilized aptamer and surface improves correct aptamer folding due to minimization of steric hindrance. (3) Positive net charge of the surface may cause the aptamer to unfold on the surface. This can be prevented by capping of amino groups with succinic acid resulting in negative surface charge. Analytical Chemistry, Vol. 80, No. 19, October 1, 2008

7377

Two different surfaces were used. While GAPS-modified surfaces result in excellent SNR values, not all aptamers may maintain functionality on this substrate. By utilization of PEImodified surfaces, aptamers with a higher steric demand can be immobilized in a functional manner. The developed method was applied successfully to three different aptamers, indicating a general applicability of the developed immobilization strategy. In order to use aptamers selected under different conditions, modifications of the protocol were necessary. To develop more complex aptamer microarrays

7378

Analytical Chemistry, Vol. 80, No. 19, October 1, 2008

that enable the multiplex detection of proteins, it would be useful to select the aptamers under same conditions. ACKNOWLEDGMENT This work was supported by DBU (Az. 13121).

Received for review May 28, 2008. Accepted July 24, 2008. AC801081V