Entrapment of Fluorescence Signaling DNA Enzymes in Sol−Gel

In general, all DNAzymes retained at least partial catalytic function when entrapped in either hydrophilic or hydrophobic silica-based materials, but ...
0 downloads 0 Views 608KB Size
Download PDF
Anal. Chem. 2007, 79, 3494-3503

Entrapment of Fluorescence Signaling DNA Enzymes in Sol-Gel-Derived Materials for Metal Ion Sensing Yutu Shen,† Gillian Mackey,† Nicholas Rupcich,†,‡ Darin Gloster,§ William Chiuman,§ Yingfu Li,*,†,§ and John D. Brennan*,†

Department of Chemistry, McMaster University, Hamilton, Ontario, Canada, L8S 4M1, and Department of Biochemistry and Biomedical Sciences, McMaster University, Hamilton, Ontario, Canada, L8N 3Z5

Three fluorescence signaling DNA enzymes (deoxyribozymes or DNAzymes) were successfully immobilized within a series of sol-gel-derived matrixes and used for sensing of various metal ions. The DNAzymes are designed such that binding of appropriate metal ions induces the formation of a catalytic site that cleaves a ribonucleotide linkage within a DNA substrate. A fluorophore (fluorescein) and a quencher (DABCYL, [4-(4dimethylaminophenylazo)benzoic acid]) were placed on the two deoxythymidines flanking the ribonucleotide to allow the generation of fluorescence upon the catalytic cleavage at the RNA linkage. In general, all DNAzymes retained at least partial catalytic function when entrapped in either hydrophilic or hydrophobic silica-based materials, but displayed slower response times and lower overall signal changes relative to solution. Interestingly, it was determined that maximum sensitivity toward metal ions was obtained when DNAzymes were entrapped into composite materials containing ∼40% of methyltrimethoxysilane (MTMS) and ∼60% tetramethoxysilane (TMOS). Highly polar materials derived from sodium silicate, diglycerylsilane, or TMOS had relatively low signal enhancements, while materials with very high levels of MTMS showed significant leaching and low signal enhancements. Entrapment into the hybrid silica material also reduced signal interferences that were related to metal-induced quenching; such interferences were a significant problem for solution-based assays and for polar materials. Extension of the solid-phase DNAzyme assay toward a multiplexed assay format for metal detection is demonstrated, and shows that sol-gel technology can provide new opportunities for the development of DNAzyme-based biosensors. Immobilization of biomolecules for sensing applications has been widely used in analytical sciences in recent decades. For example, both proteins1 and single-stranded DNA species2 have been immobilized onto a range of surfaces, typically through * To whom correspondence should be addressed. Tel: (905) 525-9140 ext 27033. Fax: (905) 527-9950. E-mail: [email protected] or [email protected]. † Department of Chemistry. ‡ Current address: Genpharm Inc., 85 Advance Rd., Etobicoke, ON, Canada, M8Z 2S9. § Department of Biochemistry and Biomedical Sciences.

3494 Analytical Chemistry, Vol. 79, No. 9, May 1, 2007

physical adsorption, covalent binding, or affinity-based interactions (i.e., avidin-biotin),3 to create analytical devices such as microarrays, affinity columns, and biosensors. However, proteins often suffer from issues related to control of orientation and denaturation during and after covalent immobilization,1 while modification of DNA is often required to conjugate functional groups to allow covalent binding to occur.2 Physisorption of either proteins or DNA, while simpler to perform, can lead to dissociation of bound biomolecules under certain pH and ionic strength conditions and may prevent proper folding of proteins or functional nucleic acids (aptamers or DNAzymes).4 Furthermore, proteins and DNA immobilized by such methods are prone to degradation by proteases5 and nucleases,6 respectively. As a result of these issues, there is need for new methods for immobilization of functional biomolecules and, in particular, functional nucleic acids. An emerging method for bioimmobilization is entrapment of biomolecules into sol-gel-derived silica materials. While widely reported as a method for immobilization of both soluble and membrane-bound proteins,7-12 reports on the entrapment of DNA into sol-gel-derived materials are few and are generally restricted to studies on the nature of nucleotide-silica interactions13 or the use of DNA as a silica templating agent.14 While it has been suggested that DNA-based biorecognition elements could potentially be entrapped with sol-gel-derived materials,8,15 there is only one report on DNA hybridization within silica.16 (1) Camarero, J. A. Biophys. Rev. Lett. 2006, 1, 1-28. (2) Henke, L.; Krull, U. J. Can. J. Anal. Sci. Spectrosc. 1999, 44, 61-70. (3) Di Giusto, D. A.; King, G. C. Top. Curr. Chem. 2006, 261, 131-168. (4) Vandenberg, E. T.; Brown, R. S.; Krull, U. J. In Immobilized Biosystems in Theory and Practical Applications; Veliky, I. A., Mclean, R. J. C., Eds., Blackie: Glasgow, UK, 1994; pp 129-231. (5) Weetall, H. H. Appl. Biochem. Biotechnol. 1993, 41, 157-188. (6) Rupcich, N.; Nutiu, R.; Li, Y.; Brennan, J. D. Anal. Chem. 2005, 77, 43004307. (7) Jin, W.; Brennan, J. D. Anal. Chim. Acta, 2002, 461, 1-36. (8) Gill, I. Chem. Mater. 2001, 13, 3404-3421. (9) Avnir, D.; Coradin, T.; Lev. O.; Livage, J. J. Mater. Chem. 2006, 16, 10131030. (10) Pierre, A. C. Biocatal. Biotransform. 2004, 22, 145-170. (11) Coradin. T.; Boissiere, M.; Livage, J. Curr. Med. Chem. 2006, 13, 99-108. (12) Besanger, T. R.; Brennan, J. D. J. Sol-Gel Sci. Technol. 2006, 40, 209-225. (13) Pierre, A.; Bonnet, J.; Vekris, A.; Portier, J. J. Mater. Sci.: Mater. Med. 2001, 12, 51-55. (14) (a) Numata, M.; Sugiyasu, K.; Hasegawa, T.; Shinkai, S. Angew. Chem., Int. Ed. 2004, 43, 3279-3283. (b) Yin, H.; Wei, Y. Polym. Mater. Sci. Eng. 2002, 87, 271-272. (15) Gill, I.; Ballesteros, A. Trends Biotechnol. 2000, 18, 282-296. 10.1021/ac070235u CCC: $37.00

© 2007 American Chemical Society Published on Web 03/23/2007

Recently, our groups reported on the first successful immobilization of a functional nucleic acid,6 specifically a structureswitching, signaling DNA aptamer,17 within sol-gel-derived materials. The entrapped DNA aptamer retained its ability to bind to its cognate ligand, adenosine triphosphate, and to undergo a structural switch to produce a fluorescence signal. Moreover, the entrapped aptamer proved to be resistant to degradation by nucleases, making it potentially useful for in vivo analyses. In a more recent study, our groups further demonstrated that the signaling DNA aptamer could be co-entrapped with a protein enzyme and the entrapped aptamer could accurately report the activity of the protein enzyme in the solid phase.18 To further explore the potential of sol-gel immobilization technology as a platform for the immobilization of functional nucleic acids, we have expanded our studies to a second class of functional nucleic acidssdeoxyribozymes. Deoxyribozymes, also known as catalytic DNA, DNA enzymes, or DNAzymes, are singlestranded DNA molecules with catalytic capabilities.19-23 A large array of deoxyribozymes have been created by in vitro selection to perform diverse chemical transformations involving RNA cleavage,24,25 RNA ligation,26,27 DNA cleavage,28 DNA ligation,29,30 DNA phosphorylation,31,32 etc. These species possess many unique properties relative to proteins, including chemical stability, the ability to withstand denaturation and renaturation cycles, cost efficiency, and ease of site-specific labeling, as well as the simplicity of the in vitro selection method for DNAzyme generation. These features have led to these species becoming desirable alternatives to traditional protein-based enzymes and ribozymes for applications such as fluorescence-based biosensing and sensing of species such as Pb(II), which can modulate the catalytic activity of a specific DNAzyme.33,34 In this work, three fluorescence signaling DNA enzymes (named OA-II, OA-III, and OA-IV) were examined after sol-gel entrapment, each of which contained a unique fluorescence (16) Li, J.; Tan, W.; Wang, K.; Yang, X.; Tang, Z.; He, X. Proc. SPIE-Int. Soc. Opt. Eng. 2001, 4414, 27-30. (17) Nutiu, R.; Li, Y. J. Am. Chem. Soc. 2003, 125, 4771-4778. (18) Rupcich, N.; Nutiu, R.; Li, Y.; Brennan, J. D. Angew. Chem., Int. Ed. 2006, 45, 3295-3299. (19) Achenbach, J. C.; Chiuman, W.; Cruz, R. P.; Li, Y. Curr. Pharm. Biotechnol. 2004, 5, 321-336. (20) Silverman, S. K. Org. Biomol. Chem. 2004, 2, 2701-2706. (21) Joyce, G. F. Annu. Rev. Biochem. 2004, 73, 791-836. (22) Breaker, R. R. Nature 2004, 432, 838-845. (23) Peracchi, A. ChemBioChem 2005, 6, 1316-1322. (24) Breaker, R. R.; Joyce, G. F. Chem. Biol. 1994, 1, 223-229. (25) Santoro, S. W.; Joyce, G. F. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 42624266. (26) Flynn-Charlebois, A.; Wang, Y.; Prior, T. K.; Rashid, I.; Hoadley, K. A.; Coppins, R. L.; Wolf, A. C.; Silverman, S. K. J. Am. Chem. Soc. 2003, 125, 2444-2454. (27) Wang, Y.; Silverman, S. K. J. Am. Chem. Soc. 2003, 125, 6880-6881. (28) Carmi, N.; Balkhi, S. R.; Breaker, R. R. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 2233-2237. (29) Sreedhara, A.; Li, Y.; Breaker, R. R. J. Am. Chem. Soc. 2004, 126, 34543460. (30) Cuenoud, B.; Szostak, J. W. Nature 1995, 375, 611-614. (31) Li, Y.; Breaker, R. R. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 2746-2751. (32) Achenbach, J. C.; Jeffries, G. A.; McManus, S. A.; Billen, L. P.; Li, Y. Biochemistry 2005, 44, 3765-3774. (33) Nutiu, R.; Mei, S.; Liu, Z.; Li, Y. Pure Appl. Chem. 2004, 76, 1547-1561. (34) (a) Liu, J.; Lu, Y. J. Am. Chem. Soc. 2003, 125, 6642-6643. (b) Liu, J.; Lu, Y. J. Fluoresc. 2004, 14, 343-354. (c) Liu, J.; Lu, Y. J. Am. Chem. Soc. 2005, 127, 12677-12683. (d) Liu, J.; Lu, Y. Methods Mol. Biol. 2006, 335, 275288, (e) Lu, Y.; Liu, J.; Li, J.; Bruesehoff, P. J.; Pavot, C. M.; Brown, A. K. Biosens. Bioelectron. 2003, 18, 529-540.

Figure 1. Schematic representation of the fluorescence signaling mechanism in the DNAzymes studies in this work. The black line and gray curve represent the fluorophore and quencher-modified RNA/ DNA substrate and DNA enzyme, respectively. Catalysis causes the single RNA linkage to be cleaved, resulting in an increase in fluorescence intensity due to dequenching.

signaling element based on a ribonucleotide flanked by a fluorophore and quencher. These DNA catalysts were derived in a previous in vitro selection experiment.35 Since these deoxyribozymes require divalent metal ion cofactors for catalytic activity, they were utilized as model metal ion sensors in the current study. Prior to catalysis, the DNAzyme is in a low fluorescence state owing to quenching of the fluorophore by the nearby quencher. Addition of specific metal ions leads to activation of the DNAzyme, and catalysis of a cleavage reaction at the ribonucleotide, which produces dequenching and hence a large increase in fluorescence intensity (see Figure 1).36-39 The signaling abilities of the deoxyribozymes were examined both in solution and in a range of solgel materials to determine an optimal material for DNAzyme entrapment. Furthermore, extension of the solid-phase DNAzyme assay toward a multiplexed assay format for metal detection is demonstrated. This work shows that sol-gel technology can be used to immobilize functional nucleic acids, in particular DNA enzymes, providing new opportunities for the development of DNAzyme-based biosensors and solid-phase assays. EXPERIMENTAL SECTION Oligonucleotides and Other Chemicals. T4 DNA ligase, T4 polynucleotide kinase (PNK), and adenosine 5′-triphosphate (ATP) were purchased from MBI Fermentas. Tetramethylorthosilicate (TMOS), methyltrimethoxysilane (MTMS), glycerol (anhydrous), and Dowex 50 × 8-100 cation-exchange resin were obtained from Sigma (Oakville, ON, Canada). Sodium silicate (SS; technical grade, 9% Na2O, 29% silica, 62% water) was purchased from Fisher Scientific (Pittsburgh, PA). Diglycerylsilane (DGS) was prepared from TMOS by methods described elsewhere.40,41 Costar clear 384-well polystyrene microtiter plates were purchased from Nalge Nunc International (Rochester, NY). Water was purified with a Milli-Q Synthesis A10 water purification system. All other chemicals and solvents used were of analytical grade and were used without further purification. (35) Chiuman, W.; Li, Y. J. Mol. Biol. 2006, 357, 748-754. (36) Mei, S. H. J.; Liu, Z.; Brennan, J. D.; Li, Y. J. Am. Chem. Soc. 2003, 125, 412-420. (37) Liu, Z.; Mei, S. H. J.; Brennan, J. D.; Li, Y. J. Am. Chem. Soc. 2003, 125, 7539-7545. (38) Shen, Y.; Brennan, J. D.; Li, Y. Biochemistry 2005, 44, 12066-12076. (39) Shen, Y.; Chiuman, W.; Brennan, J. D.; Li, Y. ChemBioChem 2006, 7, 13431348. (40) Besanger, T. R.; Chen, Y.; Deisingh, A. K.; Hodgson, R.; Jin, W.; Stanislas, M.; Brook, M. A.; Brennan, J. D. Anal. Chem. 2003, 75, 2382-2391. (41) (a) Brook, M. A.; Chen, Y.; Guo, K.; Zhang, Z.; Jin, W.; Deisingh, A.; Brennan, J. D. J. Sol-Gel Sci. Technol. 2004, 31, 343-348. (b) Brook, M. A.; Chen, Y.; Guo, K.; Zhang, Z.; Brennan, J. D. J. Mater. Chem. 2004, 14, 14691479.

Analytical Chemistry, Vol. 79, No. 9, May 1, 2007

3495

Preparation of DNA Enzymes. All standard or modified oligonucleotides were prepared by automated DNA synthesis using cyanoethylphosphoramidite chemistry (Keck Biotechnology Resource Laboratory, Yale University; Central Facility, McMaster University), purified by 10% preparative denaturing (8 M urea) polyacrylamide gel electrophoresis (PAGE), and quantified by standard spectroscopic methods as previously described.36 The cis-acting DNAzymes were produced by ligating 5′-TTTGTGTCCGTGCFRQGGTTCGATCAAG (F, fluorescein-dT; Q, DABCYL [4-(4-dimethylaminophenylazo)benzoic acid]-dT; R, adenine ribonucleotide; all other nucleotides are deoxyribonucleotides); with a relevant 5′-phosphorylated DNA oligonucleotide in the presence of a matching DNA oligonucleotide as a template. DNA phosphorylation reactions were performed in a 50 mM Tris-HCl (pH 7.6, 25 °C), 10 mM MgCl2, 5 mM DTT, 0.1 mM spermidine, 0.1 mM EDTA solution containing 0.1 unit/µL PNK and 1 mM ATP. DNA ligation reactions were carried out in a solution containing 40 mM Tris-HCl (pH 7.8, 25 °C), 10 mM MgCl2, 10 mM DTT, 0.5 mM ATP, and 0.05 unit/µL T4 DNA ligase. Fluorescein and DABCYL labels were incorporated into the oligonucleotide sequences during automated DNA synthesis using fluorescein-dT amidite and DABCYL-dT amidite (Glen Research, Sterling, VA). The adenine ribonucleotide linkage was also introduced during solid-state synthesis using A-TOM-CE Phosphoramidite (Glen Research). The TOM (triisopropylsilyloxymethyl) protecting group on the 2′-hydroxyl group of the RNA linkage was removed by incubation with 150 µL of 1 M tetrabutylammonium fluoride in THF at 60 °C with shaking for 6 h, followed by the addition of 250 µL of 100 mM Tris (pH 8.3) and further incubation with shaking for 30 min at 37 °C. The DNA was recovered using ethanol precipitation and dissolved in water containing 0.01% SDS, and the tetrabutylammonium salt was removed by centrifugation using a spin column (Nanosep 3K Omega, Pall Corp., Ann Arbor, MI). Entrapment of DNA Enzymes. A range of silica precursors were used to prepare sols for DNAzyme entrapment, including SS, DGS, TMOS, and TMOS/MTMS mixtures. Sodium silicate sols were prepared by mixing 10 mL of ddH2O with 2.9 g of SS solution (pH ∼13) followed by addition of 5 g of Dowex cationexchange resin to replace Na+ with H+. The mixture was stirred for 30 s to reach a final pH of ∼4 and then vacuum filtered through a Bu¨chner funnel. The filtrate was then further filtered through a 0.45-µm membrane syringe filter to remove any particulates in the solution. To make the DGS precursor sol, 0.5 g of solid DGS was dissolved in 1 mL of ddH2O and the resulting mixture was sonicated for 15 min. To make MTMS and TMOS sols, 1.4 mL of ddH2O and 0.1 mL of 0.1 N HCl were added to 4.5 mL of MTMS or TMOS and the resulting mixture was sonicated for 20 min in ice-cold water. In order to make MTMS-TMOS mixtures, the 4.5 mL of silane listed above was divided proportionally using volume percentages of 20-80% MTMS in TMOS and then mixed with water and acid and cohydrolyzed by sonication. In all cases, the precursor sol solutions were mixed in a 1:1 volume ratio with a buffered solution of the DNAzyme at room temperature (20 ( 1 °C), (OA-II, 50 mM HEPES, pH 6.8; OA-III, 50 mM HEPES, pH 7.5; OA-IV, 50 mM HEPES, 100 mM KCl, pH 7.5) to provide a final volume of 15 µL of material in the well of a 3496

Analytical Chemistry, Vol. 79, No. 9, May 1, 2007

384-well plate, with a final DNAzyme concentration of 250 nM. Leaching Studies. A TECAN Safire plate reader operated in bottom-read fluorescence mode was used for examination of DNAzyme signaling and leaching in solution and sol-gel materials. All fluorescence studies were done at room temperature. The DNAzyme was excited at 460 nm, and the emission intensity was recorded at 520 nm using a 5-nm excitation bandpass and a 12nm emission bandpass, with 20 excitation pulses averaged per sample. In all cases, the emission from a blank well containing the buffer or sol-gel material was also recorded, and this value was subtracted from the emission intensity of the DNAzymecontaining well to provide accurate emission intensity values. To determine the extent of leaching of the DNA enzymes after entrapment in different sol-gel-derived materials, 384-well plates containing entrapped DNAzyme samples were incubated for 1 h with 20 µL of 50 mM HEPES buffer: pH 6.8 for OA-II, pH 7.5 for OA-III, and pH 7.5 with 100 mM KCl for OA-IV. Following this period, the supernatant was removed from the wells and emission scans were taken (2-nm increment; 460-nm excitation; 500-650nm emission) for both the supernatant and the original gel using the TECAN Safire operated in bottom-read mode. Spectra of appropriate blank wells were also collected and subtracted to eliminate background emission from the microwell plate. The emission intensity was integrated over the emission wavelength range to determine the emission level. These values were used to determine the percentage of DNA enzymes that had leached from the gels, based on the relative intensity of the emission from the gel and supernatant after dilution correction of the emission intensity of the supernatant. The leaching was also calculated by determining the change in emission intensity of the DNAzyme within the gel before and after washingsin both cases there was good agreement in the calculated extent of leaching. Time-Dependent Responses of Metal Binding DNAzymes. Assays of entrapped DNAzymes were done in 384-well plates using sol-gel materials that had been previously incubated to remove leachable DNAzymes. After incubation, the incubation buffer described above was replaced with 40 µL of buffer containing specific metal ions to achieve final concentrations of divalent metals as follows: OA-II, 10 mM MgCl2 and 1 mM NiCl2; OA-III, 20 mM MnCl2; OA-IV, 10 mM MnCl2 and 1 mM CdCl2. This concentration level is referred to as “100%”, referring to the fact that these metal concentrations led to maximum cleavage of the DNAzymes as determined by PAGE.35 Upon addition of metal ions, the emission intensity was recorded over a period of at least 30 min. Signaling data were initially obtained using full emission scans (1 min/scan, 5.5-min increments over 33 min) to assess the effects of background signals; all other data were collected at a single emission wavelength of 520 nm using the parameters noted above at an increment of 10 s/point. All data were background corrected, and the solution intensity data were also corrected for dilution factors while intensity data for entrapped DNAzymes were not. Sol-Gel Optimization Studies. The OA-II, OA-III, and OAIV DNA enzymes were entrapped in various silanes (DGS, SS, TMOS and MTMS/TMOS mixtures) within the wells of a 384microwell plate. Following gelation, 20 µL of the entrapment buffer was added to the top of each gel to equilibrate the pH for at least 1 h and remove leachable DNAzymes. Following removal of the

supernatant, initial readings of the fluorescence signals of the entrapped DNAzymes (FI) and blanks (FB) were recorded using the TECAN Safire to provide a corrected initial emission intensity (FI - FB). Then, 20 µL of a buffer containing the appropriate divalent metal ions was added, so that the final concentrations of divalent metals were at the “100%” levels noted above. The changes in fluorescence intensity were recorded over the next 90 min (measured fluorescence, FM), the raw fluorescence values at each time point were corrected by blank subtraction (FM - FB), and the relative fluorescence enhancement was calculated at each time point by dividing [(FM - FB)/(FI - FB)]. Negative controls were also performed by preparing wells with appropriate DNAzymes to which the same quantities of buffer, but no divalent metals, were added. In all cases, the total fluorescence enhancement was determined from the maximum fluorescence enhancement, which was usually the final intensity reached after 90 min. However, in some cases, the intensity values reached a plateau value and subsequently showed a decrease in intensity. In such cases, the fluorescence enhancement still refers to the maximum enhancement observed but is not equivalent to the final intensity. The response times were also determined from the kinetic response curves by calculating the initial slope of the response (change in relative fluorescence vs time) over the first 3 min after addition of divalent metal ions. Concentration Sensitivity Studies. Based on the results obtained from the sol-gel optimization study, a subset of solgel materials was selected for more extensive evaluation of their metal ion sensitivity relative to solution. Specifically, sodium silicate was chosen as a representative hydrophilic sol-gel material, while a composite material composed of 40% MTMS and 60% TMOS was chosen as a representative hydrophobic material. The latter material was also chosen because it showed superior performance in terms of total signal enhancements for each DNAzyme (see below). Following gelation, 20 µL of the entrapment buffer was added to each well for 1 h and then removed to extract leachable DNAzyme, followed by addition of 20 µL of reaction buffer containing various percentages of the usual divalent metal concentration used for each DNAzyme (0-200%). The final relative signal enhancements for each DNA enzyme at each divalent metal concentration were calculated after a 90-min incubation time as described above. For solution studies, 7.5 µL of buffer, as in the previous assay, with 500 nM DNA enzyme, was added to each well in a 384-well plate. After an initial reading on the TECAN Safire, 7.5 µL of 2× reaction buffer with a range of divalent metal concentrations, as noted above, was added to each well, diluting the DNA enzymes to 250 nM, and emission intensity was recorded over time as above. The emission intensity values were corrected for dilution factors, and relative signal enhancements at each divalent metal ion concentration were determined from final intensity values, as noted above. Multiplexed Detection of Metal Ions. A 384-well plate was used to compare the fluorescence enhancements of OA-II, OAIII, and OA-IV in the presence of a number of divalent metals, which were added either singly or in pairs at concentrations that provided maximum fluorescence enhancements in previous concentration-dependence studies. Each “DNAzyme” well contained a gel formed from 7.5 µL of either a sodium silicate sol or a 40%

MTMS/60% TMOS sol and 7.5 µL of DNA enzyme or mutant in the appropriate entrapment buffer. These samples were aged for 24 h prior to testing. “Blank” wells were prepared by using wells containing only buffer. DNA enzyme and inactive mutant sample gels were equilibrated for 1 h with metal-free buffer to remove leachable DNAzymes before 20 µL of a reaction solution with various metals present was added to initiate the cleavage reaction for entrapped samples. Metals were added as follows (from top to bottom; note that 100 mM KCl is also present in all OA-IV samples): 50 mM HEPES buffer; 1 mM MnCl2 in HEPES buffer, 0.1 mM NiCl2 in HEPES buffer, 0.1 mM CdCl2 in HEPES buffer, 0.1 mM CoCl2 in HEPES buffer, and 1 mM MnCl2 + 0.1 mM CdCl2 in HEPES buffer. A Typhoon Fluorimager was used to obtain a fluorescence image of the reaction array 1 h after the addition of metals to the wells. The fluorescence intensity values from each well was first corrected by subtracting an appropriate blank, and then the signal enhancements for each pair of replicate wells were calculated from the image obtained after metal ion addition by calculating the ratio of the intensity in the sample wells (those with active DNAzymes) to the mutant wells (i.e., wells containing mutant inactive DNAzyme to which metal ions were added). Note that OA-III intensity values were corrected using the OA-IV mutant. This method of data analysis was chosen to avoid the need to obtain two images for each analysis, so that only a single image would be needed to generate all test and control data. The intensity from mutant wells with added metals was compared to that of mutants with buffer to assess the degree of metal-induced quenching for each DNAzyme. RESULTS AND DISCUSSION Overview of DNA Enzymes Utilized. A total of three DNAzymes were selected for entrapment into sol-gel-derived materials. These catalytic DNA species were selected as model metal ion sensing DNAzymes to allow us to probe issues with detection of cationic species and assess changes in detection sensitivity, response times, quenching of fluorescence by the divalent metal ions42 (which could lead to nonlinear responses), leaching of DNAzymes, and metal ion selectivity upon entrapment of DNAzymes into different sol-gel-derived materials. All of the metal-sensitive DNAzymes were previously isolated in our laboratories by in vitro selection and are designated as OA-II, OA-III, and OA-IV.35 Each of the three deoxyribozymes exhibits distinct profiles of divalent metal ion specificities. PAGE studies previously showed that OA-II was active in the presence of Mg(II) and Ni(II), OA-III strictly used Mn(II), while OA-IV required both Mn(II) and Cd(II) for optimal activity.35 The secondary structures of these DNAzymes, determined in our previous report,35 are shown in Figure 2. The sequences of all DNA molecules used in this study are given in Table 1. Fluorescence Assay Development. The use of microwell plate-based assays required significant care to remove issues related to background signals. In general, plastic microwell plates have much higher levels of background fluorescence emission relative to cuvettes and often show different emission intensities from different wells. After evaluating several types of plates, it was determined that the lowest background signals were obtained (42) Rupcich, N.; Chiuman, W.; Nutiu, R.; Mei, S.; Flora, K. K.; Li, Y.; Brennan, J. D. J. Am. Chem. Soc. 2006, 128, 780-790.

Analytical Chemistry, Vol. 79, No. 9, May 1, 2007

3497

Figure 2. Putative secondary structures of cis-acting metal sensing deoxyribozymes: OA-II, OA-III, and OA-IV. The black nucleotides show the deoxyribozyme; gray nucleotides denote the substrate; F is fluorescein-dT; Q is Dabcyl-dT; rA is the RNA linkage.

Figure 3. Fluorescence spectra for OA-II in a 384-microwell plate as a function of time after addition of metals. Bottom spectrum is for OA-II in metal-free buffer. Other spectra are obtained 5.5, 11, 16.5, 22, 27.5, and 33 min after addition of 10 mM MgCl2 and 1 mM NiCl2. Inset shows the changes in relative emission intensity for DNAzymes in the presence of buffer and metal ions. All spectra are collected with an excitation wavelength of 460 nm, excitation bandpass of 5 nm, and emission bandpass of 12 nm.

from Costar clear 384-well polystyrene microtiter plates. Examination of top- and bottom-read fluorescence methods showed that while background signals were lower using the top-read method, the signal-to-background ratio was much better when using bottom-read fluorescence measurements. The signal-to-background level was also improved by reducing the excitation bandpass, increasing the emission bandpass, and blue-shifting the excitation wavelength away from the absorbance maximum to minimize backscatter of excitation light into the detector. Optimal settings for the TECAN plate reader (which utilizes monochromators rather than filters for wavelength selection) were 460-nm excitation (5-nm bandpass) and 520-nm emission (12-nm bandpass) with integration of 20 excitation pulses to reduce noise in the sample and background signals. Even with these optimized conditions, the background signal contributed significantly to the initial fluorescence signal (prior to addition of divalent metal ions), to the point where ∼80% of the signal was due to background (data not shown). However, these conditions provided a sufficient difference in initial fluorescence and background levels to produce good reproducibility between samples (80% leaching. The results suggest that the inclusion of high levels of MTMS leads to a more open pore structure and that DNA-silica interactions do not aid in retaining the biomolecule. The degree of leaching is relatively high compared to proteins44 and is likely a result of the electrostatic repulsion of DNA from the silica surface, which could prevent silica templating around the DNA, and thus result in relatively high mobility for the entrapped biomolecule. Overall, the data above show that composite materials formed from 40% MTMS/60% TMOS provide the best overall performance for entrapped DNAzymes, with the highest signal enhancements, (43) Flora, K.; Brennan, J. D. Analyst 1999, 124, 1455-1462. (44) Zheng, L.; Reid, W. R.; Brennan, J. D. Anal. Chem. 1997, 69, 3940-3949.

Figure 6. Fluorescence enhancements for DNA enzymes in solution and sol-gel materials as a function of divalent metal ion concentration. The signals are the final fluorescence readings, relative to the initial signal after equilibration. The 100% divalent metal concentration is 10 mM Mg2+ and 1 mM Ni2+ for OA-II, 20 mM Mn2+ for OA-III, and 10 mM Mn2+ and 1 mM Cd2+ for OA-IV. The error bars represent the standard deviation between the two replicate wells. (A) Response of DNAzymes in solution. (B) Average effect of metal ions on F-DNA quenching. (C) Response of DNAzymes in 40% MTMS. (D) Response of DNAzymes in sodium silicate.

highest signaling rates, and lowest degree of leaching. Such materials have been previously used to entrap hydrophobic proteins such as lipases45 and, in some cases, can also be very useful for entrapment of more polar proteins.46 It is likely that such materials have a balance of hydrophobic and hydrophilic character that allows for retention of DNAzymes within the pores while also allowing facile entry of divalent metal ions with minimal metal-silica interactions. More polar materials are likely to interact with the metals, while more hydrophobic materials could prevent entry of cations into the matrix. Concentration-Dependence Studies. To further assess the effects of entrapment in hydrophobic and hydrophilic materials, and to better understand the reasons for the unusual responses obtained in polar materials (viz. Figure 4), signaling levels were examined as a function of metal ion concentration of the OA-IIOA-IV enzymes in solution and when entrapped in sodium silicate or 40% MTMS/60% TMOS-derived materials (Figure 6). In solution assays, the optimal metal concentrations to obtain maximum signal enhancements were generally 1-10% of the amount of metal (45) (a) Reetz, M. T.; Zonta, A.; Simpelkamp, J. Biotechnol. Bioeng. 1996, 49, 527-534. (b) Reetz, M. T.; Zonta, A.; Simpelkamp, J.; Konen, W. Chem. Commun. 1996, 1397-1398. (c) Reetz, M. T.; Zonta, A.; Simpelkamp, J. Agnew. Chem., Int. Ed. Engl. 1995, 34, 301-303. (d) Kuncova, G.; Guglielmi, M.; Dubina, P.; Safar, B. Collect. Czech. Chem. Commun. 1995, 60, 15731577. (46) (a) Brennan. J. D.; Hartman, J. S.; Ilnicki, E. I.; Rakic, M. Chem. Mater. 1999, 11, 1853-1864. (b) Gulcev, D.; Goring, G. L. G.; Rakic, M.; Brennan. J. D. Anal. Chim. Acta 2002, 457, 47-59. (c) Goring, G. L. G.; Brennan, J. D. J. Mater. Chem. 2002, 12, 3400-3406.

needed to obtain the maximum rate of cleavage of the DNAzyme (Figure 6A).35 Beyond this level, the fluorescence signal decreased, which was assumed to be due to quenching of fluorescence by the metal ions.42 To test this hypothesis, each of the metal ion solutions was added over the same concentration range to the F-DNA fragment that would be generated upon cleavage of the DNAzymes (note: this fragment is identical in OA-II-IV). The addition of metals led to similar decreases in emission intensity (average quenching is shown in Figure 6B), which has previously been shown to be due to a combination of static and dynamic quenching of the fluorophore by metal ions in solution and bound to the anionic DNA backbone.42 However, the metalinduced quenching occurred at very low metal ion concentrations (0-10%) relative to the range where the DNAzymes showed significant decreases in intensity (usually between 5 and 100%). Thus, dividing the relative enhancement of the DNAzyme by the quenching of the F-DNA at each metal concentration did not lead to a suitable correction, suggesting that the use of F-DNA was not appropriate as a control. These data indicate that other mechanisms may be at play to cause the decreases in intensity, such as reassociation of F-DNA and Q-DNA strands. The data suggest that alternative controls are needed to develop a multiplexed assay format in order to achieve useful data from metalsensing DNAzymes. Note that this may not be a significant problem for DNAzymes that detect other species such as organic molecules or proteins, which do not produce significant fluorescence quenching. Analytical Chemistry, Vol. 79, No. 9, May 1, 2007

3501

Figure 7. (A) Fluorimages of a 384-microwell plate taken after the addition of divalent metal ions. The final concentration of the DNAzyme in the gel was 250 nM. (B) Relative signal enhancements upon addition of various metals to DNAzymes and mutants, showing responses obtained from MTMS and SS samples side by side for each DNAzyme in the presence of various metals.

The change in signal with metal ion concentration was somewhat altered upon entrapment, with moderate quenching occurring for 40% MTMS (Figure 6C) and significant quenching occurring in sodium silicate-based materials (Figure 6D) for all three DNAzymes tested. Interestingly, for SS, the maximum signal was reached in the presence of 5-10% of the optimal metal ion concentration, which is similar to the case observed in solution. Furthermore, the maximum signal enhancements were relatively high (3-6-fold) as compared to the value obtained at the 100% metal concentration (1.3-2-fold), with maximum signal levels occurring in the 100 µM-1 mM range and detection limits being in the low-micromolar range. However, beyond the concentration of metal ion corresponding to the maximum signal, all three DNAzymes demonstrated decreases in fluorescence, indicating a need to have control samples tested to account for quenchinginduced changes in signal. In the case of 40% MTMS-derived materials, the concentration-response curve is somewhat broader (particularly for OA-II and OA-III), and quenching is significantly lowered, suggesting that capping of the anionic silica surface leads to a decrease in metal preconcentration and decreased quenching. Only the OA-IV system shows a response similar to solution, and even here, the quenching is suppressed to a significant degree relative to SS and solution samples. Multiplexed Detection of Metal Ions. A key issue with the detection of metal ions is the competition between signal enhancement due to cleavage of the ribo linkage and the inherent quenching of fluorescence by the metal ions. As shown in Figure 6, the nonlinear change in signal with metal ion concentration 3502

Analytical Chemistry, Vol. 79, No. 9, May 1, 2007

makes it difficult to perform metal analysis. Furthermore, the fact that the three DNAzymes show some cross-selectivity could potentially make identification of specific metal ions somewhat difficult. In an attempt to overcome these problems, we designed a multiplexed DNAzyme array in a 384-well plate format that contained each of the three active DNAzymes as well as specific mutant sequences (see Table 1) to act as a quenching control. Mutant sequences were used instead of F-DNA as controls for the sol-gel studies owing to significant leaching of the small F-DNA species from sol-gel materials. The mutant sequences do not have the ability to be cleaved, have the same sequence as the active DNAzymes in the region of the fluorescent label, and thus should show similar leaching and quenching behavior. The active DNAzymes and mutant species were entrapped in both 40% MTMS/60% TMOS and sodium silicate gels, along with appropriate blanks, and imaged after addition of a series of metal ions. Figure 7A shows the Fluorimage of the plates after metal ion addition. Before the metals were added to the plate, all of the DNA-containing samples showed similar levels of fluorescence while the blanks showed no detectable fluorescence signals (image not shown). However, after addition of metals the active DNAzyme samples exhibited the expected patterns of fluorescence enhancement based on previous PAGE studies.35 As shown in Figure 7A, OA-II was sensitive primarily to Ni(II); OA-III was most sensitive to the presence of Mn(II), either when alone or combined with Cd(II), while OA-IV required both Mn(II) and Cd(II) for maximum signal enhancement, regardless of the type of sol-gel material used. On the basis of raw intensity values from

the image, it is clear that the samples based on 40% MTMS are brighter than those derived from SS, although the increased brightness is also obtained for the mutant samples in 40% MTMS relative to SS. Figure 7B shows quantitative data derived from the array image obtained after addition of metal ions. The data show that the enhancements range from 2.5- to 4-fold for 40% MTMS samples, and from ∼2- to 3-fold for SS samples. These enhancements are lower than those shown in Figure 4 owing to the use of the DNAzyme/mutant signals for the array instead of the ratio of signals before and after metal addition, as done in Figure 4. In general, the MTMS samples showed similar or higher enhancements than SS samples, as expected, and the use of the mutant species allowed for on-array correction that produced the expected pattern of response. The mutant species showed only minor changes in fluorescence intensity in the presence of different metal ions, often undergoing only slight enhancements in intensity in the presence of metals (vs buffer), suggesting that the metals have minimal direct effects on the signal at the levels used in this study. Overall, the multiplexed data analysis highlights the ability to detect multiple metals either alone or in mixtures and suggests that there is potential for moving to a microarray format for multiplexed analysis using sol-gel-entrapped DNAzymes.47 CONCLUSIONS This work is the first example of the entrapment of deoxyribozymes within sol-gel-derived materials. The data clearly show that such materials have significant versatility for the entrapment of a range of viable fluorescence signaling DNAzymes, extending the potential applications of such materials. Indeed, the same materials used to entrap DNAzymes have previously been shown to be amenable to entrapment of proteins,45,46 suggesting that both species could be coimmobilized in the same matrix. Interestingly, it was determined that maximum sensitivity toward metal ions (47) Rupcich, N.; Goldstein, A.; Brennan, J. D. Chem. Mater. 2003, 15, 18031811. (48) Hodgson, R.; Chen, Y.; Zhang, Z.; Tleugabulova, D.; Long, H.; Zhao, X.; Organ, M. G.; Brook, M. A.; Brennan, J. D. Anal. Chem. 2004, 76, 27802790. (49) Rimmele, M. ChemBioChem. 2003, 4, 963-971.

was obtained when DNAzymes were entrapped into composite materials containing ∼40% MTMS and 60% TMOS. Highly polar materials derived from sodium silicate, diglycerylsilane or TMOS had relatively low signal enhancements, while materials with very high levels of MTMS showed significant leaching and low signal enhancements. Importantly, such composite materials have also been previously shown to provide a barrier to entry of enzymes that could degrade entrapped functional nucleic acids,6 to provide higher levels of biomolecule loading that is possible with monolayer coatings,47 and are amenable to many platforms, including microarrays, thin-film coatings, or bioaffinity columns.48 Aptamers, DNAzymes, and allosteric DNAzymes provide a useful addition to the range of available biorecognition elements, since they can be selected for virtually any biological target, from small metabolites to large proteins, and in conjunction with a signal production method, such as that used herein, provide a tool for rapid diagnostic analysis of complex biological samples. Furthermore, their properties can be manipulated easily, since the strands are created synthetically and thus can be derivatized in a number of ways to alter signaling, binding, or catalytic function. A particular advantage of aptamers is their ability to bind strongly to small metabolites, an area where elicitation of monoclonal antibodies is difficult.49 The use of immobilized DNAzymes for sensing and diagnostic applications will be discussed in future publications. ACKNOWLEDGMENT The authors thank the Natural Sciences and Engineering Research Council of Canada for funding this work through a network grantsSENTINEL Canadian Network for the Development and Use of Bioactive Paper. The authors also thank the Canadian Institutes for Health Research, the Ontario Ministry of Energy, Science and Technology, the Canada Foundation for Innovation, the Ontario Innovation Trust, and MDS-Sciex for support of this work. Y.L. is a Canada Research Chair in Nucleic Acids Biochemistry. J.D.B. holds the Canada Research Chair in Bioanalytical Chemistry. Received for review February 4, 2007. Accepted February 20, 2007. AC070235U

Analytical Chemistry, Vol. 79, No. 9, May 1, 2007

3503