Anal. Chem. 2005, 77, 5460-5466
Aptamer-Derived Nucleic Acid Oligos: Applications to Develop Nucleic Acid Chips to Analyze Proteins and Small Ligands Rika Yamamoto-Fujita† and Penmetcha K. R. Kumar*
Institute for Biological Resources and Functions, National Institute of Advanced Industrial Science and Technology (AIST), Central 6, 1-1 Higashi, Tsukuba Science City 305-8566, Japan
The specificity and affinity of aptamers for their cognate ligands are comparable to those of antibodies for antigens. To use aptamers effectively in high-throughput assays in a microarray format, to analyze various analytes, we developed a strategy in which the aptamer was split into two nonfunctional units and allowed to reassemble into the functional aptamer by the cognate ligand. We have named this method “analyte-dependent oligonucleotide modulation assay” (ADONMA). As proof-of-principle, we used oligonucleotides derived from the aptamer RNA against HIV-1 Tat and demonstrated, with both titer plates and plastic slide chips, that specifically in the presence of Tat or its peptide, the two oligos reconstituted the core binding regions of Tat. Thus, these results suggest that ADNOMA has the potential for use in nucleic acid microarrays for detecting various ligands. Several high-affinity RNA and DNA aptamers have been isolated from combinatorial libraries of nucleic acids that bind to a wide variety of targets, ranging from simple ions to whole cells.1 The affinities displayed by the aptamers (nucleic acids) are comparable to the affinities of antibodies for antigens; therefore, their utility as a fundamental molecular recognition element in biosensors has been realized.1-3 To date, diverse analytical formats have been reported for detecting various analytes by aptamers.4-7 An enzyme-linked oligonucleotide assay (ELONA) was developed to detect human vascular endothelial growth factor (hVEGF) levels in sera.7 The ELONA yielded results equivalent to those from an enzyme-linked immunosorbent assay (ELISA), with similar accuracy, specificity, and interference. Therefore, it appears that the in vitro-evolved aptamers could potentially substitute for antibody use in clinical research and diagnostics. Although the aforemen* Corresponding author. Phone: 81-29-861-6085. Fax: 81-298-61-6159. Email:
[email protected]. † Present address: Research Institute for Cell Engineering, National Institute of Advanced Industrial Science and Technology (AIST), Kansai Center, Ikeda, 1-8-31 Midorigaoka, Ikeda-shi, Osaka 563-8577, Japan. (1) Osborn, S. E.; Ellington, A. D. Chem. Rev. 1997, 97, 349-370. (2) O’Sullivan, C. K. Anal. Bioanal. Chem. 2002, 372, 44-48. (3) Clark, S. L.; Remcho, V. T Electrophoresis 2002, 23, 1335-1340. (4) Kleinjung, F.; Klussmann, S.; Erdmann, V. A.; Scheller, F. W.; Furste, J. P.; Bier, F. F. Anal. Chem. 1998, 70, 328-331. (5) Potyrailo, R. A.; Conrad, R. C.; Ellington, A. D.; Hieftje, G. M. Anal. Chem. 1998, 70, 3419-3425. (6) Lee, M.; Walt, D. R. Anal. Chem. 2000, 282, 142-146. (7) Drolet, D. W.; Moon-McDermott, L.; Roming, T. S. Nat. Biotechnol. 1996, 14, 1021-1025.
5460 Analytical Chemistry, Vol. 77, No. 17, September 1, 2005
tioned studies have been encouraging and promising,4-7 in these studies, full-length aptamers were employed. This approach may pose several limitations: for example, the efficiency of chemical synthesis diminishes with the increasing length of the aptamer; the full-length aptamer must be modified for protection against nucleases; and in the case of multiple analyses, the folding and refolding are relatively inefficient for longer aptamers. To reduce the overall cost for synthesis and the aforementioned limitations to use full-length aptamers in biosensors, we previously showed that molecular beacon aptamers have the ability to detect the Tat protein in solution.8 In these studies, the fulllength RNA aptamer sequence was separated into two strands, and the fluorophore and quencher were attached at the 5′- and 3′-ends, respectively, of the RNA strand that forms a hairpin structure. Specifically, in the presence of Tat or its peptides, but not in the presence of other RNA binding proteins, the two oligomers underwent a conformational change to form a duplex that relieved the fluorophores from the quencher, and thus, a significant enhancement of fluorescein fluoroscence was observed. In recent years, among the various multiplex and highthroughput assays, microarray technology has been growing most rapidly. These immobilized nucleic acid arrays have proven useful for the rapid detection of mutations and polymorphisms, as well as for discovery and expression monitoring. At present, these arrays are limited to analyzing only complementary nucleic acids and cannot be used for other analytes. To expand the applications of nucleic acid arrays for analyzing various ligands, including proteins and small ligands, such as metabolites and nucleotides, we have explored the use of aptamer-derived oligos. In this study, we show that the aptamer-derived oligos can be redesigned and adapted to a microarray analysis format to detect ligands quantitatively. MATERIALS AND METHODS RNA, Tat-Peptide, and Tat Protein. The Tat-1-derived peptide, CQ (aa 37-72) was chemically synthesized, deprotected, and purified by HPLC. The amino acid composition of the synthetic peptide was confirmed after cleaving each peptide with 6 N HCl containing 0.1% phenol at 16 °C for 1 h, followed by HPLC analysis. The aptamer RNATat oligonucleotides reported in this manuscript were all synthesized on an RNA/DNA synthesizer (Applied (8) Yamamoto, R.; Baba, T.; Kumar, P. K. R. Genes Cells 2000, 5, 389-396. 10.1021/ac050364f CCC: $30.25
© 2005 American Chemical Society Published on Web 07/29/2005
Figure 1. Secondary structures of TAR, aptamer RNATat-derived oligo RNAs, and gel-shift assay for these modulating oligos: (A) The TAR-1 and TAR-2 RNAs and the aptamer RNATat; shadowed and outlined letters indicate core elements that are required for Tat binding. (B) Modulating aptamer RNAs (i, DA-1/DA-2; ii, DA-3/DA-4; iii, DA-5/DA-6; iv, DA-7/DA-8; v, DA-1/DA-4). (C) Representative autoradiograms from gel-shift assays for various modulating aptamers: lane 1, radiolabeled 5′-oligo (10 nM); lane 2, radiolabeled 5′-oligo (10 nM) and unlabeled 3′-oligo (200 nM); lane 3, radiolabeled 5′-oligo (10 nM) and unlabeled 3′-oligo (200 nM) in the presence of CQ (20 nM); lane 4, radiolabeled 5′-oligo (10 nM) and unlabeled 3′-oligo (200 nM) in the presence of 200 nM Tat-1. Shadowed and bold arrows indicate the positions of the ternary complexes. The free 5′-oligonucleotide is indicated by a thin arrow.
Biosystems, model 394) using phosphoroamidites from Glen Research (Glen Research Corporation, Sterling, VA). The aptamer RNAs were deprotected and purified as described previously.9 The 5′-strands of the RNAs were labeled with γ32P-ATP using T4 polynucleotide kinase.9 Similarly, the 3′-fluoresceinated oligomer RNA (3′-fluorescein DA-9), the 3′-biotin-labeled DA-10 RNA, the A11 and A13 oligos with an NH2 modification at the 5′end with a C6 spacer length, and the A12 and A14 oligos with a Cy3 fluorophore at the 5′end were also synthesized chemically and deprotected using the protocols recommended by the supplier. Gel-Shift Assay. Ternary complex formation in the presence of RNA oligonucleotides was assayed using a previously reported (9) Yamamoto, R.; Katahira, M.; Nishikawa, S.; Taira, K.; Kumar, P. K. R. Genes Cells 2000, 5, 371-388.
protocol5 in the presence of either Tat or the Tat peptide CQ. Eight RNA oligonucleotides with the potential to form five duplexes having different duplex stability were prepared and characterized (Figure 1). In all cases, the 5′-strand of the aptamerderived oligo RNAs was labeled with γ32P-ATP (DA-1, DA-3, DA5, DA-7, DA-5i, and DT-1). The 5′-end labeled RNA (2 Kcpm) was mixed with 200 nM of unlabeled complementary RNA (DA-2 for DA-1, DA-4 for DA-3, DA-1 for DA-4, DA-6 for DA-5, DA-8 for DA7, DA-6i for DA-5i, and DT-2 for DT-1 oligo) in the presence of 40 nM unlabeled Escherichia coli tRNA in 10 µL of Tat binding buffer (10 mM Tris-HCl, pH 7.8; 70 mM NaCl; 2 mM EDTA; and 0.01% Nonidet P-40). The CQ peptide (20 nM) or the Tat-1 protein (200 nM) was added, and the reaction mixture was incubated at 30 °C for 1 h. The reaction products were separated on a nondenaturing Analytical Chemistry, Vol. 77, No. 17, September 1, 2005
5461
polyacrylamide gel (15%), and the amounts of complex formed in the presence and absence of protein or peptides were quantified by an image analyzer (BAS2000, Fuji Film, Japan). Equilibrium dissociation constants (Kd) were determined for the aptamer-derived oligo RNAs for complexes DA-1/DA-2 and DA-5/DA-6 in the presence of the CQ peptide. The 5′-end-labeled RNA (50 pM) DA-1 or DA-5 was mixed with the appropriate complementary RNA in 10 µL of Tat binding buffer, and 40 nM of tRNA was added as a nonspecific competitor. The CQ peptide (0.5-64 nM) was added, and the incubation proceeded at 30 °C for 1 h. The free and complex products were separated on a nondenaturing polyacrylamide gel (15%) and quantified. Bmax and Kd were determined by saturation radioligand binding data, which were fit to the following binding equation,
Y ) BmaxX/Kd + X
where Y is the specific binding, Bmax is the maximum binding, and X is the concentration of the ligand. A nonlinear regression analysis was carried out using the Graphpad PRISM software (Graphpad Software Inc., San Diego, CA). Analyte-Dependent OligoNucleotide Modulation Assay (ADONMA) Using Titer Plates. Streptavidin-coated microtiter plates were purchased from Labsystems, Finland. The 3′-biotinylated oligonucleotide DA-10 (5 pmol) was added to the well and allowed to bind with streptavidin in 50 µL of Tat binding buffer at ambient temperature for 10 min, followed by washing with 200 µL of Tat binding buffer. Tat binding buffer (50 µL) containing 10 pmol of 3′-fluoresceinated DA-9 and either 200 pmol of Tat or various concentrations of CQ was then added. The microtiter plate was incubated at 30 °C for 30 min and then washed with 200 µL of Tat binding buffer. The fluorescence was quantified using a fluoroimager, with excitation at 488 nm and detection at 530 nm. Controls were carried out in the absence of Tat-1 protein or Tatderived peptides. Assays were also carried out using 8 units of HeLa nuclear extract (Promega, U.S.A.) in the presence of 40 units of RNase inhibitor (Toyobo, Japan). ADONMA Using Plastic Slides. Two kinds of surface-coated plastic slides (size 76 × 26 mm) were obtained from Sumitomo Bakelite, Japan: an aldehyde surface and a Sumitomo proprietary surface “S-Bio PrimeSurface”, referred to here as A and B, respectively. To spot the aptamer-derived oligos (A11 and A13) onto these slides, 1 and 10 µM concentrations of oligos were prepared in spotting buffer (100 mM Tris buffer pH7.4, 700 mM NaCl, and 20 mM EDTA). Each concentration of aptamer was spotted 48 times on the slide using a spotting machine (“Chot”, Nichiryo, Japan) with a pin diameter of 500 µm. The coated slides were baked for 1 h at 80 °C. To remove the unbound oligos from the slides, the surface A and B slides were incubated in 0.25% NaBH4, PBS/EtOH (4:1 ratio v/v), and 0.1 N NaOH, respectively, for 5 min. The slides were then washed for 2 min each with boiling water and cold water, and dried after an ethanol wash. To observe the specific ternary complex, 50 µL of a sample containing CQ (5 µM, final concentration) and the hybridizing oliog A12 or A14 were applied to the respective slides (for the anchor oligo A11-coated slide, the A12 oligo with the CQ peptide was applied). The sample was spread over the whole slide by applying the cover slip, and 5462
Analytical Chemistry, Vol. 77, No. 17, September 1, 2005
the slide was incubated for 5 min. The unbound hybridizing oligo and CQ were removed by three washes with 200 mL of diluted buffer (1 mM Tris-HCl, pH 7.4; 7 mM NaCl; and 0.2 mM EDTA). The slides were dried using a compressed air spray, and the spots were quantitated by a microarray scanner (ScannArray Lite, PerkinElmer Life Sciences, Boston, MA). RESULTS AND DISCUSSION Several aptamer-derived oligonucleotides of different lengths, which include the core binding elements for the Tat-1 protein, were synthesized (core elements boxed in Figure 1A). Biochemical studies previously identified a minimal recognition region of Tat-1 that is sufficient for RNA recognition and binding. The peptide region comprising amino acid residues 37-72, known as the CQ peptide, binds to the TAR-1 RNA (59-mer) and aptamer RNA as efficiently as the full-length Tat-1 protein,9-10 and therefore, we used CQ instead of Tat-1 in our binding kinetics experiments, mainly to avoid possible folding problems with the Tat-1 protein during expression and purification. Screening for Efficient Oligos Derived from an Aptamer That Binds to the Tat-Peptide. To design suitable aptamerderived oligos that form a stable and specific ternary complex (by virtue of forming a stable duplex structure) only in the presence of the cognate target, we have prepared five aptamer-derived oligos and tested them using a gel-shift assay (Figure 1, B and C, i-v). The oligos contains all the important bases necessary for binding to the Tat protein (indicated by the bold letters), but they are differ in stem lengths upon forming the duplex and, thus, have different duplex stabilities ranging from -7.1 to -15.2 kcal/mol, evaluated by the RNA draw program 3.71. Four of these, DA-1/ DA-2, DA-3/DA-4, DA-5/DA-6, and DA-1/DA-4 (i, ii, iii, and v, respectively), formed specific complexes (Figure 1C) in the presence of Tat protein (200 nM, lanes 4) or Tat-derived peptide (CQ, 20 nM, lanes 3), but not in the absence of Tat protein (lanes 2). These results suggests that the above oligos reconstitute the Tat binding pocket by virtue of forming a duplex structure and are therefore useful in monitoring indirectly the amount of Tat present in the samples. Among various oligos tested, the DA-5/ DA-6 oligonucleotides formed a complex more efficiently than the other oligos in the presence of the Tat protein (50%) or the CQ peptide (84%) (Figure 1C, iii vs i, ii, and v). However, increasing further the stem length by one additional base pair at one end of the duplex, the aptamer-derived oligos appear to form a duplex even in the absence of Tat at room temperature in the above binding buffer, and thus, such oligos may not be useful for monitoring Tat in the infected samples (data not shown). The Kd values for the DA-5/DA-6-CQ and DA-1/DA-2-CQ ternary complexes were analyzed by gel-shift assays in the presence of various concentrations of CQ (0.1 to 12.8 nM and 2 to 64 nM, respectively). The Kd values for the DA-5/DA-6-CQ and DA-1/DA-2-CQ complexes were 0.5 and 16 nM, respectively (Figure 2). The DA-5/DA-6-CQ complex showed an affinity 32fold higher than that of the DA-1/DA-2-CQ complex. The difference between these two complexes lies mainly in the formation of two additional G-C base pairs in the complex, at both ends of the molecule. These results suggest that the oligos derived from (10) Matsugami, A.; Kobayashi, S.; Ouhashi, K.; Uesugi, S.; Yamamoto, R.; Taira, K.; Nishikawa, S.; Kumar, P. K. R.; Katahira, M. Structure 2003, 11, 533545.
Figure 2. Determination of the equilibrium dissociation constants (Kd) for the binding of aptamer-derived oligo RNAs, DA-1/DA-2 and DA-5/DA-6, to CQ. Representative autoradiograms from a gel-shift assay, showing binding analyses of the aptamer-derived RNAs DA1/DA-2 and DA-5/DA-6 to CQ (A and B, respectively). Bold and thin arrows indicate the positions of the ternary complex (RNA-CQ) and the free 5′-oligonucleotide, respectively. Saturation curves and Scatchard plots are shown for the DA-1/DA-2-CQ and the DA-5/DA-6CQ compelxes (A and B, respectively).
the aptamer are reconstituted as an RNA duplex in the presence of Tat or the CQ peptide, and the specificity and affinity of this duplex for the Tat protein are comparable to those of the hairpin aptamerTat (Figure 1A, left). Importance of Sequence and Functional Groups of Oligos Derived from the Aptamer. Since the Kd for the DA-5/DA-6CQ complex was estimated to be close to the value for the hairpin aptamer RNATat (Figure 1A, left), it is likely that the DA-5/DA-6 oligos reconstitute the binding core elements in the presence of the Tat protein or the CQ peptide. The core binding elements of the aptamer RNATat for binding to the Tat protein consist of a central 4-base-pair helix flanked by two bulges containing two residues each. In addition, site-specific mutagenesis studies suggested that the U residues at both bulges are very important for binding to the Tat protein. On the basis of kinetic binding studies of the hairpin aptamer and the DA-5/DA-6 aptamer oligonucleotides, it appears that the functional groups near the bulge region are important for specific interaction with the Tat protein. To confirm this notion, two derivatives of DA-5 and DA-6 were synthesized with C-to-U substitutions and analyzed for
Figure 3. Gel-shift analysis of the ability of the aptamer-derived oligos DA-5/DA-6, the inactive aptamer-derived oligos, and the TAR RNA-derived oligonucleotides to bindr Tat-derived peptides. (A) Secondary structures of inactive aptamer-derived oligo variants of RNA (i, DA-5i/DA-6i) and duplex TAR RNA (ii, DT-1/DT-2). (B) Representative autoradiograms from gel-shift assays for the binding of inactive aptamer-derived oligos and TAR RNA-derived oligonucleotides to CQ (i and ii, respectively): lane 1, radiolabeled 5′-oligo (10 nM); lane 2, radiolabeled 5′-oligo and unlabeled 3′-oligo (200 nM); lane 3, radiolabeled 5′-oligo and unlabeled 3′-oligo (200 nM) in the presence of CQ (20 nM).
forming complex as mentioned. The resulting oligonucleotide pair (DA-5i/DA-6i) failed to form a ternary complex (Figure 3A,i and B,i), thus suggesting that important bases identified earlier9 are, indeed, essential for the formation of the complex in these aptamer-derived oligos. Oligonucleotides derived from the aptamer were also subjected to a gel-shift binding assay in the presence of other RNA binding proteins, such as the HCV NS3 protein (which has a protease domain and an RNA helicase domain). The NS3 protein failed to form a complex with these oligonucleotides, as revealed by the gel-shift analysis (data not shown). The binding of CQ to DA-5/ DA-6 was also compared to another set of RNA oligonucleotides derived from the TAR-1 RNA, DT-1/DT-2. Figure 3A,ii and 3B,ii clearly shows that DT-1/DT-2 did not form a ternary complex, even in the presence of a large excess of oligonucleotide (DT-2) and CQ peptide. Taken together, these results suggest that the oligos derived from the aptamer are able to reconstitute efficiently the binding pocket and stabilized the duplex structure by the Tat protein or its peptides compared to the authentic TAR RNA-1. ADONMA Using Titer Plates. The aptamer-derived oligos DA-5/DA-6 described above have potential as a diagnostic tool for the detection of the Tat protein. One such diagnostic assay was developed and tested and is presented in Figure 4. This assay, which we refer as analyte-dependent oligonucleotide modulation assay (ADONMA), uses a 3′-fluorescein oligonucleotide (DA-9), a 3′-biotin oligonucleotide (DA-10) and streptavidin-coated microtiter plates (Figure 4A and B). Before analyzing these oligos (DA-9 Analytical Chemistry, Vol. 77, No. 17, September 1, 2005
5463
Figure 4. ADONMA using titer plates. (A) The aptamer-derived oligos DA-9 and DA-10 used in ADONMA. (B) Schematic representation of ADONMA. (C) Concentration dependence of ADONMA using CQ peptide: intensity of the fluorescent signal in the absence of Tat-derived peptide, CQ (control), or in the presence of CQ (10, 50, and 100 pmol). (D) ADONMA using Tat-1: control, in the absence of Tat-1 or CQ; 1, in the presence of CQ (10 pmol; orange bar); 2, in the presence of Tat-1 (200 pmol; red bar). (E) The effect of HeLa nuclear extract on ADONMA: intensity of fluorescent signal in the absence of CQ (control); in the presence of 8 units of HeLa nuclear extract, 1 (green bar); or in the presence of CQ (10 pmol; orange bar), 2. The relative fluorescence intensities shown in panels C, D, and E are from three independent experiments (SEM indicated by error bars). The concentrations of the fluorescent and biotin-labeled RNA oligonucleotides and other experimental details are in the Methods Section.
and DA-10) in titer plates, we compared DA-9/DA-10 oligos’ with DA-5/DA-6 oligos’ ability to form a ternary complex with the CQ peptide in native gels and found that these oligos have nearly the same equilibrium dissociation constant (0.5 nM). Initially, the biotinylated DA-10 oligonucleotide (5 pmol) was added to the streptavidin plates, and the unbound oligonucleotide was removed by washing. The fluorescein-modified DA-9 oligonucleotide (10 pmol) was then added with the Tat-1 or CQ peptide, and a final washing step was performed to remove the unbound material. Appropriate controls were carried out in the absence of Tat-1 or CQ peptide. The fluorescence intensity was quantified using a fluoroimage analyzer, and the results are summarized in Figure 4C. In the presence of increasing amounts of CQ peptide (10100 pmol), the fluorescence increased about 500- to 2500-fold, as compared to the control without CQ. A higher concentration of Tat-1 protein (200 pmol) was required to achieve a fluorescence signal comparable to that of the CQ peptide (10 pmol; Figure 4D). To be an effective diagnostic tool, the assay shown in Figure 3A should provide quantitative results in the presence of crude samples, such as a mammalian cell nuclear extract. The crude samples might contain proteins and compounds that interfere with the assay as well as proteins that promote the annealing of complementary sequences.11 The assay described above was carried out with DA-9/DA-10, a HeLa nuclear extract, and an RNase inhibitor. Figure 4E shows that the modulating aptamer oligonucleotides do not assemble into the duplex and are not retained in the microtiter well in the presence of the HeLa nuclear (11) Portman, D. S.; Dreyfuss, G. EMBO J. 1994, 13, 213-221.
5464
Analytical Chemistry, Vol. 77, No. 17, September 1, 2005
Figure 5. Sequences of the anchoring and hybridizing oligos are A11, A13, and A12, A14, respectively. The red letter indicates the position of the substitution, and C6 indicates the spacer length between the NH2 and the first base of the oligo.
extract alone (it should be noted here that a small amount of RNase activity was detected, despite the presence of 40 units of RNase inhibitor). ADONMA Using Plastic Chip. To evaluate the ability of the aptamer-derived oligos (DA-5/DA-6), when coated on a plastic chip, to detect their cognate target, we prepared two kinds of aptamer-derived oligos (A11/A12 and A13/A14, based on DA-5/ DA-6 oligos), and analyzed them on two different plastic chip
Figure 6. ADONMA using surface A plastic slides. (A) Scan view of surface A with the A11/A12 oligos in either the presence or absence of the RE peptide. (B) Scan view of surface A with the A13/A14 oligos in either the presence or absence of the RE peptide. (C) Quantitative analysis of intensities of spots (averaged from 48 spots).
Figure 7. ADONMA using surface B plastic slides. (A) Scan view of surface B with the A11/A12 oligos in either the presence or absence of the RE peptide. (B) Scan view of surface B with the A13/A14 oligos in either the presence or absence of the RE peptide. (C) Quantitative analysis of spot intensities (averaged from 48 spots).
surfaces (aldehyde, surface A, and a proprietary original coat, “SBio PrimeSurface” surface B) available from Sumitomo Bakelite, Japan. These oligos were chemically synthesized and analyzed, initially for the formation of a ternary complex on the native gels. As expected, the A11/A12 oligos formed a ternary complex only in the presence of the Tat-derived peptide, CQ, but in their absence, no detectable amount of duplex formation was observed (data not shown). The DA-11/DA-12 oligos have an affinity comparable to that of DA-5/DA-6 oligos. On the other hand, the A13/A14 oligos, which have lethal mutations that affect their ability to bind to Tat, were affected significantly in terms of forming either a ternary or duplex structure in the absence or presence of the Tat-derived peptides. Next, the anchoring oligos, A11 and A13 with an NH2 group at the 5′-end, and hybridizing oligos A12 and A14, having a Cy3 (cyanine derivative dye) fluorophore at the 5′end, are prepared (Figure 5). To obtain a suitable signal for detection and analysis, we spotted two concentrations of the A11and A13 oligos (1 and 10 µM) on the chips. On each plastic slide, we coated 48 spots at 1 and 10 µM concentrations. The hybridizing oligos, A12 and A14, were incubated in either the presence or the absence of the Tat-derived peptide (CQ, 5 µM final concentrations). The uncomplexed oligos were washed away, and the amount of fluorescence retained on the spot was analyzed using a microarray scanner (ScanArray Lite). The results clearly show that in the presence of the Tat-derived peptide, CQ, the signal-to-noise ratio was ∼6-fold higher (Figures 6 and 7). However, in the absence of CQ peptide, the signal-to-noise ratio was nearly unchanged. Among the two surfaces of the slides tested, surface B was found to be the most suitable for this kind of analysis because the hybridizing oligos specifically hybridized where the anchor oligo was spotted (Figure 7). Surface A
(aldehyde surface) appeared to interact nonspecifically with the hybridizing oligo and to spread all over the chip (Figure 6). This could not be avoided, even after including several washing steps. The mutant oligos (A13/A14), as expected, did not show any significant increase in the fluorescence in the presence of the Tatpeptide, suggesting the importance of the Tat-binding region within these oligos (Figures 6 and 7). However, a similar mutant aptamer completely lost the binding ability to the CQ peptide in the native gel analysis when the analysis was carried out at subnanomolar concentrations. The observed difference between the two assays might have originated from the higher concentrations (micromolar) used in the plastic chip assay. Although the oligos used in the present studies are sensitive to ribonucleases, these oligos can be readily converted into ribonuclease resistant by incorporating a 2′-fluoro modification of pyrimidines. Such modified oligos have shown an affinity to the Tat of HIV (unpublished data) that is simiar to that of unmodified DA5/DA-6 oligos. Taken together, these results as well as the titer plate analysis clearly indicate that the aptamer-derived oligos have potential applications in detecting analytes, such as proteins. CONCLUSION The use of aptamer-derived oligos to detect specific analytes, as described here, has many advantages over methods that use the full-length aptamer sequence, including the following: (1) shorter RNA oligonucleotides can be synthesized with higher efficiency than longer molecules, (2) no modification of the analyte is required, (3) modifications to stabilize the nucleic acid are necessary for only a portion of the aptamer, (4) proper folding of the modulating aptamer is facilitated by the analyte, and (5) lower cost. The results presented above suggest that the modulating Analytical Chemistry, Vol. 77, No. 17, September 1, 2005
5465
aptamer method has potential for detecting analytes of interest. Moreover, further improvements will be achieved when the RNA aptamers can be fully protected from ribonucleases. Even though the method described here is for a protein that recognizes the bulge regions between the two stems, it can be expanded to proteins or small molecules that recognize stemiloops or other alternative RNA structures. This approach can be generalized for
5466
Analytical Chemistry, Vol. 77, No. 17, September 1, 2005
the detection of proteins other than HIV Tat, by selecting appropriate nonmodulating and/or modulating species from combinatorial libraries. This approach is currently underway for several viral proteins as well as for small molecules. Received for review March 2, 2005. Accepted July 2, 2005. AC050364F
