DNA-programmed chemistry in rapid homogeneous assays for DNA and protein targets.

DNA-programmed chemistry (DPC) is a novel technology for synthesis of a wide variety of organic compounds at nanomolar concentrations under physiologic conditions (1 ). Annealing of small molecule-oligonucleotide precursors to DNA templates, each with an attached chemical moiety, can generate high amounts of effective molarities of these reactants. This phenomenon can enhance, by nearly 1-million-fold, the rate of a reaction, while increasing its specificity. We have developed a specific chemical process that generates a fluorescent product on annealing of 2 oligonucleotides to each other. Oligonucleotides containing a 3 -terminal azidocoumarin (AzC) and a 5 -terminal triphenylphosphine (TPP) will react in aqueous solution to form a fluorescent product, 7-aminocoumarin. The reaction rate is very slow at submicromolar concentrations of the single-stranded TPP and AzC groups in free solution. However, if these fluorophore precursor-containing oligonucleotides anneal to each other or to a common DNA target, as long as the 3 and 5 fluorophore precursors are annealed in close proximity, their localized high concentration supports their reaction to yield a fluorescent product. We have found that this principle can also be exploited for detection of proteins (and other high molecular weight analytes), with the advantages of a simple, homogeneous phase assay with potentially very low background. We modified this architecture for detecting proteins and other non–nucleic-acid targets. The model target selected was platelet-derived growth factor (PDGF-BB), a protein that contains 2 identical B subunits, for which tight-binding aptamers have been previously identified. We obtained human PDGF-BB and mouse monoclonal antihuman PDGF-BB from R&D systems. Except as indicated, all reaction and melting curve solutions contained, in a volume of 100 L: 50 mmol/L Tris/HCl buffer at pH 8.0, 10 mmol/L magnesium chloride, 40 pmol of detection oligonucleotides, 40 pmol of PDGF-BB, and typically, 350 mL/L formamide by volume. We determined melting curves by measuring fluorescence enhancement of SYBR Green I (Molecular Probes) dye binding to double-stranded DNA in a Bio-Rad iCycler under conditions similar to those described by Lipsky et al. (2 ). Melting curves were obtained in a 1000-fold dilution of SYBR Green dye, increasing temperature from 10 °C to 90 °C in 0.5 °C increments every 10 s. Melting temperature (Tm) was defined as the point of inflection of the first derivative of the melting curve, according to iCycler software. For fluorescence generation by TPP and AzC, the increase of fluorescence with time was monitored by incubation at 25 °C in a Wallac Victor 1420 luminometer, with excitation at 355 nm and emission at 460 nm. We prepared TPP oligonucleotides from resin-bound 5 -amino oligonucleotides converted to their TPP derivatives (3 ) and azidocoumarin oligonucleotides by reaction of 3-amino oligonucleotides with the N-hydroxylsuccinimide ester of 7-azio-4-methylcoumarin-3-acetic acid (4 ). Oligonucleotides were purified by C18 chromatography. Within the oligonucleotide sequences, the anti-PDGF-B aptamer sequence (5 ) is underlined and reporter sequences are in italics. The spacer sequence between the aptamer and reporter sequences in all cases was CCCCCCCCCC. All sequences are indicated 5 to 3 . Coumarin aptamer oligonucleotide: CAG GCT ACG GCA CGT AGA GCA TCA CCA TGA TCC TGC CCC CCC CCC ATA TTT AAG C-AzC; TPP “matched” oligonucleotide: TPP-GCT TAA ATA TCC CCC CCC CCC AGG CTA CGG CAC GTA GAG CAT CAC CAT GAT CCT G; TPP “mismatched” oligonucleotide: TPP-TGG GAA TGG TGC CCC CCC CCC CAG GCT ACG GAC GTA GAG CAT CAC CAT-GAT CCT G. We synthesized 3 oligonucleotides for this assay. The oligonucleotide containing the 3 -AzC group contained the 5 PDGF-BB aptamer sequence, a C10 spacer, and a 10-base reporter sequence with a 3 terminal AzC group. The TPP oligonucleotide contained the 3 -PDGF-BB aptamer sequence, a C10 spacer, and a 10-base 5 reporter sequence (complementary to the reporter sequence of the AzC oligonucleotide) containing the 5 TPP group. The 3rd oligonucleotide had the same architecture as the 2nd oligonucleotide, but the reporter sequence was mismatched to the AzC oligonucleotide. Under experimental conditions of pH 8, 25 °C, 10 mmol/L magnesium chloride, and 350 mL formamide, the complementary reporter sequences of the oligonucleotides (measured Tm 20 °C in 35% formamide) did not form a stable duplex (determined by SYBR Green melting curve analysis), and in the DPC-based assay, the fluorophore precursors on the oligonucleotide probes reacted very slowly in the absence of a target (Fig. 1). When we added the target for the aptamers, PDGF-BB, the fluorescent product 7-aminocoumarin was generated, producing a fluorescent signal by the 2 aptamer-containing AzC and TPP oligonucleotide reporter molecules. Substitution of an aptamer-TPP oligonucleotide, in which the reporter sequences were mismatched to the AzC oligonucleotide, abolished the reaction (Fig. 1), demonstrating the requirement that the reporter regions of the oligonucleotides be complementary to generate fluorescence. Omission of either of the oligonucleotides in the pair or the target prevented the reaction. Titration of the 2 matched oligonucleotides with each other revealed that the optimal ratio of AzCto TPP-containing aptamer oligonucleotides was 1:1, and the optimal ratio of total oligonucleotides to PDGF-BB was 2:1, consistent with a mechanism in which the 2 different aptamer-containing oligonucleotides bind to a single molecule of PDGF-BB with 2 binding sites. (However, only half of the ternary complexes, on average, Abstracts of Oak Ridge Posters