ARTICLE pubs.acs.org/ac
Surface Immobilization of Structure-Switching DNA Aptamers on Macroporous Sol-Gel-Derived Films for Solid-Phase Biosensing Applications Carmen Carrasquilla,† Yingfu Li,†,‡ and John D. Brennan*,†,‡ † ‡
Department of Biochemistry and Biomedical Sciences, McMaster University, Hamilton, Ontario, Canada, L8N 3Z5 Department of Chemistry and Chemical Biology, McMaster University, Hamilton, Ontario, Canada, L8S 4M1
bS Supporting Information ABSTRACT: Structure-switching signaling aptamers (ss-aptamers) are single-stranded DNA molecules that are generated through in vitro selection and have the ability to switch between a duplex composed of a quencher-labeled DNA strand (QDNA) hybridized adjacent to a fluorophore label on the aptamer, and an aptamer-target complex wherein the QDNA strand is released, generating a fluorescence signal. While such species have recently emerged as promising biological recognition and signaling elements, very little has been done to evaluate their potential for solid-phase assays. In this study, we demonstrate that high surface area, sol-gel-derived macroporous silica films are suitable platforms for high-density affinity-based immobilization of functional ssaptamer molecules, allowing for binding of both large and small target analytes with robust signal development. These films are formed using a poly(ethylene glycol) (PEG)-doped sodium silicate material, and we show that it is possible to control the pore size distribution and surface area of the silica film by varying the amount of PEG. Materials with the highest surface area are shown to be able to immobilize up to 6-fold more ss-aptamer than planar glass surfaces, providing greater detection sensitivity and somewhat improved detection limits as compared to immobilization on conventional glass. The solid-phase assay is performed using two different structure-switching signaling aptamers with high selectivity for adenosine 50 -triphosphate and platelet-derived growth factor, respectively, demonstrating that this immobilization scheme should be suitable for a variety of target ligands.
A
ptamers are single-stranded nucleic acids that can function as receptors for target molecules due to their ability to fold into distinct three-dimensional structures. In recent years, detection methods based on DNA or RNA aptamers have emerged owing to the versatility of these molecules as biological recognition elements. This is because aptamers can be generated by in vitro selection for virtually any target of interest1-3 and indeed have already been selected for a number of diverse analytes, including proteins, small metabolites, and cells.1,4 Moreover, the stability and ease of modification and labeling offered by these species make them ideal candidates for use in signaling biosensors.5 For example, aptamer-based fluorescent reporters that function by switching conformation from a DNA/DNA duplex to a DNA/ target complex have been developed.6 This conformational change can be coupled to a fluorescence-dequenching mechanism such that the fluorescence signal generation is dependent on the concentration of the target ligand. The employment of structure-switching aptamers (ssaptamers) for solid-phase biosensing applications relies on their immobilization on or within a suitable surface. 7-10 Typical r 2011 American Chemical Society
methods used to immobilize single-stranded DNA are based on surface attachment via covalent, electrostatic, or affinity interactions.11-17 Physical adsorption can result in the DNA molecules being immobilized in a random orientation, which may reduce activity, increase dissociation from the surface, or make them inaccessible to their target.18-20 Although the use of covalent or affinity interactions offers greater control over orientation,20,21 these methods may still suffer from poor sensitivity because of the low binding-site density afforded by the monolayer surface coverage on conventional planar surfaces. When using planar surfaces, maximal aptamer loading is restricted by low surface area and electrostatic repulsion, while augmented packing density may result in steric effects that reduce aptamer activity and signal generation,22 limiting the capabilities of sensing schemes using two-dimensional surfaces. This can be particularly problematic when using structure-switching aptamers, Received: October 10, 2010 Accepted: December 12, 2010 Published: January 7, 2011 957
dx.doi.org/10.1021/ac102679r | Anal. Chem. 2011, 83, 957–965
Analytical Chemistry as these species require the ability to undergo significant conformational changes upon target binding, which can be restricted when using high density monolayers. 23 An alternative route for bioimmobilization involves entrapment of aptamers or DNA enzymes within three-dimensional matrixes, such as polymer microspheres, hydrogels, or porous inorganic materials,14,16,24-30 that can accommodate a greater number of molecules as compared to two-dimensional immobilization techniques, improving biosensor detection limits and sensitivity. Using these methods, biomolecules are not bound to the support surface but are retained through size-exclusion, while small analyte molecules may diffuse into the region containing the biosensing species. Unfortunately, entrapment within polymeric matrixes can lead to inaccessibility to large analytes,27 while potentially harsh chemical processing conditions or propensity for some polymers to swell in aqueous solutions may lead to degradation or leaching of the biosensing species, respectively.31-33 While entrapment within sol-gel materials may not be suitable for detection of high molecular weight analytes, it is possible to immobilize biomolecules onto the surface of sol-gel-derived silica materials via covalent or affinity interactions to produce a solid-phase biosensor. 29 An advantage of the sol-gel method is the ability to tune the pore size distribution of the material through addition of porogens such as poly(ethylene glycol) (PEG). 34-36 The presence of larger pores in the network may produce a very high accessible surface area that could accommodate large amounts of DNA aptamer while retaining accessibility to larger molecular weight analytes and structure-switching ability, thereby making this method useful for the detection of proteins and even cells. In this paper, we describe the use of macroporous sol-gelderived films as high surface area supports for immobilization of structure-switching DNA aptamers in order to develop high sensitivity biosensors for the model analyte adenosine 5 0 -triphosphate (ATP, MW = 573 Da) and the homodimeric protein platelet-derived growth factor-BB (PDGF-BB, MW = 24 800 Da). In this work, thick films formed on the bottom of 96-well plates were utilized to provide a simple means for quantifying fluorescence signals from immobilized aptamers. However, the materials used to form films were selected from those with long gelation times (>30 min), since these are amenable to dipcoating onto surfaces to produce thin films that allow facile interfacing to analytical devices such as electrodes and optical fibres. 37-40 Following the formation of films, the silica surface was functionalized with a biotinstreptavidin bridge to bind biotinylated and fluoresceinlabeled DNA structure-switching aptamers to the surface. These were then hybridized to cDNA strands containing a quencher moiety (QDNA) to reduce the fluorescence intensity of the labeled aptamer. Introduction of target analytes to this surface induces a conformational change that causes dehybridization of the quencher strand that produces a large fluorescence signal (Figure 1). 6 We show that immobilization of ss-aptamers onto high surface area solgel-derived materials provides higher overall sensitivity and better structure-switching behavior relative to aptamers on planar glass surfaces, suggesting that even at high surface density the aptamers retain the necessary conformational flexibility for target binding and structure switching.
ARTICLE
Figure 1. Schematic illustration of the immobilization and action of biotinylated structure-switching aptamers on sol-gel-derived materials in microwell plates. An amine-functionalized macroporous sol-gelderived silica film is further functionalized with a biotin-streptavidin bridge. This can bind and immobilize the biotinylated bipartite construct of a structure-switching aptamer for specific target molecule detection (F = fluorophore and Q = quencher). Upon analyte binding to the aptamer complex (ATP or PDGF), the quencher strand is released resulting in a fluorescence signal enhancement. Target molecules include ATP (low MW molecule) and PDGF-BB (high MW biomolecule).
’ EXPERIMENTAL SECTION Chemicals. Modified oligonucleotides were chemically synthesized by Integrated DNA Technologies (Coralville, IA) and purified by HPLC. FITC-streptavidin was purchased from Zymed (San Francisco, CA). Streptavidin, bovine serum albumin (BSA), 10 kDa poly(ethylene glycol) (PEG), 3-(aminopropyl)triethoxysilane (APTES), biotin, N-(3-(dimethylamino)propyl)N0 -ethylcarbodiimide hydrochloride (EDC), N,N0 -disuccinimidyl carbonate (DSC), and Dowex 508-100 cation exchange resin were obtained from Sigma-Aldrich (Oakville, ON). Adenosine 50 -triphosphate (ATP), cytosine 50 -triphosphate (CTP), guanosine 50 -triphosphate (GTP), and uridine 50 -triphosphate (UTP) were purchased from Fermentas Life Sciences (Burlington, ON). Recombinant platelet-derived growth 958
dx.doi.org/10.1021/ac102679r |Anal. Chem. 2011, 83, 957–965
Analytical Chemistry
ARTICLE
factor-BB (PDGF-BB), epidermal growth factor (EGF), and insulin-like growth factor I (IGF-I) were obtained from R and D Systems (Minneapolis, MN). NeutrAvidin, FITC-NeutrAvidin, and sodium silicate solution (SS solution, ultrapure grade, ∼14% Na2O, ∼29% silica) were purchased from Fisher Scientific (Pittsburgh, PA). Glass-bottom 96-well plates were purchased from Greiner (Frickenhausen, Germany). Microwells were soaked with 100 μL of 0.1 N NaOH for 18-24 h, followed by copious washing with water and drying under a stream of nitrogen, to clean and hydrolyze the glass prior to addition of sol-gel based materials. Water was purified with a Milli-Q Synthesis A10 water purification system. All other chemicals and solvents were of analytical grade and were used as received. Procedures. Preparation of Sol-Gel-Derived Materials. Buffer solutions were prepared by mixing an equal volume of buffer (100 mM HEPES 3 NaOH; pH 6.9-7.6) with a solution containing either 0.0, 5.0, or 10.0% (w/v) 10 kDa PEG in water. Sodium silicate (SS) solutions were prepared as described elsewhere by diluting 2.6 g of a stock SS solution to 10 mL with water, mixing the solution with 5.5 g DOWEX to bring the pH of the SS solution to ∼4, and then filtering this solution through a Buchner funnel to remove the Dowex resin followed by further filtration through a 0.45-μM membrane syringe filter to remove any particulates in the solution.27 This solution was combined with the PEG-doped buffer solution in a 1:1 (v/v) ratio, and 50 μL of the sol was deposited in the cleaned wells, allowed to gel, and then aged for 4-6 h to create a thick film on the glass bottom. From the 24 materials prepared, three were selected for our studies: (1) SS-LRG (sodium silicate material, large pore size), prepared at pH 6.9 with a final concentration of 2.5% (w/v) PEG; (2) SS-MED (sodium silicate material, medium pore size), prepared at pH 7.6 with a final concentration of 5% (w/v) PEG; (3) SS-SML (sodium silicate material, small pore size), prepared at pH 7.3 without PEG. Characterization of Sol-Gel-Derived Silica Morphology. Thin monoliths of sol-gel-derived silica were prepared for porosity measurements using both nitrogen adsorption and mercury intrusion porosimetry. After gelation, all monoliths were cured in air for 4-6 h at 20 °C prior to addition of 50 mM HEPES 3 NaOH buffer and aging for 4 days with daily buffer exchanges, which was replaced by water for an additional 3 days (one exchange per day) to remove any buffer salts. Following aging and washing of silica monoliths, the samples were desiccated for 5 days. Before analysis, dried monoliths were crushed and outgassed for 8-12 h to remove air and residual water from the surface. Nitrogen porosimetry and mercury intrusion analyses were carried out as described in detail elsewhere.34 Samples for SEM imaging were aged for 5 days before analysis and coated with 5 nm of platinum under vacuum to improve conductivity. Imaging was performed at 2 kV using a JEOL JSM 7000F scanning electron microscope. Functionalization of Sol-Gel-Derived Materials. For aminosilanization, silica materials and glass surfaces were incubated with 100 μL of a solution containing 4% APTES in 650 mM acetic acid with shaking for 18 h and then washed with ddH2O to remove residual APTES. Biotin was conjugated to the surfaces via succinimide ester coupling using 100 μL of a solution containing 30 mM EDC, 60 mM DSC, and 2.4 mM biotin in 100 mM MES buffer (pH 6.8) with shaking for 12 h. Samples were again washed with ddH2O to remove any excess biotin. Modified avidin proteins, including streptavidin (SA), FITC-labeled streptavidin (FITC-SA), neutravidin (NA), or FITC-labeled neutravidin
(FITC-NA), were incubated over the glass or silica film layers in the wells at 500 nM for 18 h followed by washing with buffer to remove unbound protein. All functionalization steps were characterized by obtaining solid-state 13C cross-polarization-magic angle spinning (CP/ MAS) NMR spectra of dried monoliths using a Bruker AVANCE 300 NMR spectrometer with a Bruker standard bore MAS probe with 4-mm zirconia rotors at a frequency of 75.48 MHz, with proton decoupling during acquisition and referencing to external glycine. Samples were spun at 5 kHz using a 2 ms contact time and a 5 s delay between pulses. At least 800 scans were obtained with a window of 22.5 kHz. For comparison of loading density using tethered and entrapped biomolecules, films containing entrapped FITC-SA or FITC-NA were also prepared by mixing a SS sol with an aqueous solution containing 1 μM FITC-SA/NA in 100 mM HEPES buffer (pH 7.3), in a 1:1 (v/v) ratio followed by deposition of 50 μL of this sol into the wells to create a protein-doped silica material on the glass-bottom well. These samples were allowed to age for at least 4 h prior to addition of buffer to prevent dehydration of the silica material. Preparation of DNA Aptamer Complexes. Bipartite structure-switching aptamers for ATP or PDGF detection were prepared using the following sequences. The F designates the location of a covalently bound fluorescein label (F) in the construct. The italicized nucleotides designate the quencherDNA-binding region. ATP bipartite aptamer (ATP-FDNA): 50 -biotin-TTTTTTTTTTFTCACTGACCTGGGGGAGTATT-GCGGAGGAAGGT-30 . ATP aptamer quencher-DNA (ATP-QDNA): 50 -CCCAGGTCAGTG-dabcyl-30 . PDGF bipartite aptamer (PDGF-FDNA): 50 -FCAGGCTACGGCACGTAGAGCATCACCATGATC-CTG(T20)-biotin-30 . PDGF aptamer quencher-DNA1 (PDGF-QDNA1): 50 -CGTGCCGTAGCCTG-dabcyl-30 . PDGF aptamer quencher-DNA2 (PDGF-QDNA2): 50 -GTGCCGTAGCCTG-dabcyl-30 . PDGF aptamer quencher-DNA3 (PDGF-QDNA3): 50 -TGCCGTAGCCTG-dabcyl-30 . PDGF aptamer quencher-DNA4 (PDGF-QDNA4): 50 -GCCGTAGCCTG-dabcyl-30 . To create the bipartite complex, the fluorescently labeled aptamer strand (FDNA) and the quencher strand (QDNA) were combined in a 1:3 molar ratio in 20 mM Tris buffer at pH 7.8 (containing 100 mM NaCl and 5 mM MgCl2 for the ATP aptamer or 140 mM NaCl, 5 mM KCl, 1 mM CaCl2, and 1 mM MgCl2 for the PDGF aptamer) and incubated for at least 1 h at 25 or 37 °C, for the ATP or PDGF aptamer, respectively, to ensure annealing and quenching. For solution-based studies utilizing NA-DNA or SA-DNA complexes, a final concentration of 400 nM NA or SA was added to the solution of FDNA (100 nM final concentration) prior to incubation with a 300 nM QDNA. Immobilization of DNA Aptamers. For avidin-functionalized sol-gel-derived films and glass surfaces, 50 μL of 100 nM of the appropriate biotinylated-FDNA was incubated in buffer over the materials for 12 h at 25 °C with shaking to immobilize the DNA aptamers onto these surfaces. For the aptamer-entrapped films, FQDNA complexes were incubated with SA in buffer prior to combining with the sodium silicate solution in a 1:1 volume ratio, and 50 μL of this sol was deposited in the glass-bottom wells before being allowed to gel and age for at least 4 h. Final concentrations 959
dx.doi.org/10.1021/ac102679r |Anal. Chem. 2011, 83, 957–965
Analytical Chemistry
ARTICLE
of the biomolecules entrapped in the sol-gel material were 100 nM FDNA with 400 nM SA and 300 nM QDNA. Films were washed with the appropriate aptamer buffer prior to testing. Fluorescence Intensity Measurements. Fluorescence images were acquired using a Typhoon multimode imaging scanner with excitation at 488 nm, medium sensitivity, and 100 μm resolution. Fluorescence measurements of aptamer response and selectivity in solution and on film surfaces in microwell plates were carried out using a Tecan M1000 platereader. Solution-based thermal denaturation curves (increasing temperature from 20-90 °C at a rate of 1 °C/min) were performed using a Cary Eclipse fluorimeter. Both solution and immobilized aptamer samples were excited at 490 nm, and emission was collected at 520 nm using a 5-nm bandpass for both excitation and emission with a 0.5 s integration time on the bottom-read setting. These settings have previously been shown to provide high signal to background levels while minimizing photobleaching of the fluorescein dye;28 emission signals remained constant for periods of over an hour using these settings. Solid-phase samples were also corrected for light scattering by blank subtraction of signals originating from the materials without immobilized DNA. 28 All ATP aptamer studies were performed at 25 °C, while those involving the PDGF aptamer were performed at 37 °C. All fluorescence measurements are reported as F/F o where F is the end point fluorescence intensity and F o is the initial fluorescence intensity prior to target addition (thermal denaturation profiles are measured in relative fluorescence units, RFU). Regeneration Studies. Following initial signal enhancement measurements of the ATP aptamer immobilized on various materials, wells were shaken for 10 min in various wash solutions followed by shaking in ATP aptamer buffer for 10 min and regeneration in new buffer. Regeneration was performed by adding 300 nM QDNA to washed materials and incubating for 1 h at 25 °C prior to the addition of 1 mM ATP and retesting. The regeneration solutions tested were: 1 ATP aptamer buffer, 1 M NaCl, 10% SDS in 1 buffer, 50:50 (v:v) MeOH:water, 6 M urea, or 2.5 mM HCl.
compositions were examined in order to create macroporous films of varying porosity with long gelation times that were suitable for dipcasting (>30 min). Initial studies revealed that macroporous materials could be formed using 2-8% (w/v) of 10 kDa PEG in a SS sol containing 6 wt % of silica when using a pH of 6.9. Flocculation occurred at concentrations of less than 2%, while concentrations above 8% resulted in no phase separation, demonstrating a similar trend, though over a more narrow range of polymer concentrations, as compared to DGS materials. By screening several combinations of PEG concentration and pH, three materials were identified for further study. These included the following: (1) materials prepared at pH 6.9 and 2.5% PEG, which contained almost 50% macropore volume and a surface area of ∼47 m2/g, designated as “SS-LRG”; (2) a material prepared at pH 7.6 with 5% PEG, which contained ∼18% macropore volume and a surface area of 90 m2/g, designated as “SS-MED”, and a material formed at pH 7.3 without PEG, which had
