DNA Oligonucleotide-Templated Nanocrystals: Synthesis and Novel Label-Free Protein Detection

Abstract

Semiconductor and magnetic nanoparticles hold unique optical and magnetic properties, and great promise for bio-imaging and therapeutic applications. As part of their stable synthesis, the nanocrystal surfaces are usually capped by long chain organic moieties such as trioctylphosphine oxide. This capping serves two purposes: it saturates dangling bonds at the exposed crystalline lattice, and it prevents irreversible aggregation by stabilizing the colloid through entropic repulsion. These nanocrystals can be rendered water-soluble by either ligand exchange or overcoating, which hampers their widespread use in biological imaging and biomedical therapeutics. Here, we report a novel scheme of synthesizing fluorescent PbS and magnetic Fe3O4 nanoparticles using DNA oligonucleotides. Our method of PbS synthesis includes addition of Na2S to the mixture solution of DNA sequence and Pb acetate (at a fixed molar ratio of DNA/S2−/Pb2+ of 1:2:4) in a standard TAE buffer at room temperature in the open air. In the case of Fe3O4 particle synthesis, ferric and ferrous chloride were mixed with DNA in DI water at a molar ratio of DNA/Fe2+/Fe3+ = 1:4:8 and the particles were formed via reductive precipitation, induced by increasing pH to ∼11 with addition of ammonium hydroxide. These nanocrystals are highly stable and water-soluble immediately after the synthesis, due to DNA termination. We examined the surface chemistry between oligonucleotides and nanocrystals using FTIR spectroscopy, and found that the different chemical moieties of nucleobases passivate the particle surface. Strong coordination of primary amine and carbonyl groups provides the chemical and colloidal stabilities, leading to high particle yields (Figure 1). The resulting PbS nanocrystals have a distribution of 3–6 nm in diameter, while a broader size distribution is observed with Fe3O4 nanoparticles as shown in Figure 1b and c, respectively. A similar observation was reported with the pH change-induced Fe3O4 particles of a bimodal size distribution where superparamagnetic and ferrimagnetic magnetites co-exist. In spite of the differences, FTIR measurements suggest that the chemical nature of the oligonucleotide stabilization in this case is identical to the PbS system. As a particular application, we demonstrate that aptamer-capped PbS QD can detect a target protein based on selective charge transfer, since the oligonucleotide-templated synthesis can also serve the additional purpose of providing selective binding to a molecular target. Here, we use thrombin and a thrombin-binding aptamer as a model system. These QD have diameters of 3∼6 nm and fluoresce around 1050 nm. We find that a DNA aptamer can passivate near IR fluorescent PbS nanocrystals, rendering them water-soluble and stable against aggregation, and retain the secondary conformation needed to selectively bind to its target, thrombin, as shown in Figure 2. Importantly, we find that when the aptamer-functionalized nanoparticles binds to its target (only the target), there is a highly systematic and selective quenching of the PL, even in high concentrations of interfering proteins as shown in Figure 3a and b. Thrombin is detected within one minute with a detection limit of ∼1 nM. This PL quenching is attributed to charge transfer from functional groups on the protein to the nanocrystals. A charge transfer can suppress optical transition mechanisms as we observe a significant decrease in QD absorption with target addition (Figure 3c). Here, we rule out other possibilities including Forster resonance energy transfer (FRET) and particle aggregation, because thrombin absorb only in the UV, and we did not observe any significant change in the diffusion coefficient of the particles with the target analyte, respectively. The charge transfer-induced photobleaching of QD and carbon nanotubes was observed with amine groups, Ru-based complexes, and azobenzene compounds. This selective detection of an unlabeled protein is distinct from previously reported schemes utilizing electrochemistry, absorption, and FRET. In this scheme, the target detection by a unique, direct PL transduction is observed even in the presence of high background concentrations of interfering negatively or positively charged proteins. This mechanism is the first to selectively modulate the QD PL directly, enabling new types of label free assays and detection schemes. This direct optical transduction is possible due to oligonucleotidetemplated surface passivation and molecular recognition. This chemistry may lead to more nanoparticle-based optical and magnetic probes that can be activated in a highly chemoselective manner.