Colorimetric Detection of Mercuric Ion (Hg²⁺) in Aqueous Media using DNA‐Functionalized Gold Nanoparticles

Color is everything: Hg2+ in aqueous media is detected by the formation of thymidine–Hg2+–thymidine coordination complexes, which raises the melting temperature of the DNA-hybridized gold nanoparticle probes and thus the temperature at which the probes disperse and effect a purple-to-red color change. The method has very high sensitivity and selectivity, and it provides a simple and fast colorimetric readout (see picture). Mercury is a widespread pollutant with distinct toxicological profiles, and it exists in a variety of different forms (metallic, ionic, and as part of organic and inorganic salts and complexes). Solvated mercuric ion (Hg2+), one of the most stable inorganic forms of mercury,1 is a caustic and carcinogenic material with high cellular toxicity.2 The most common organic source of mercury, methyl mercury, can accumulate in the human body through the food chain and cause serious and permanent damage to the brain with both acute and chronic toxicity.3–5 Methyl mercury is generated by microbial biomethylation in aquatic sediments from water-soluble mercuric ion (Hg2+).4 Therefore, routine detection of Hg2+ is central to the environmental monitoring of rivers and larger bodies of water and for evaluating the safety of aquatically derived food supplies.5, 6 Several methods for the detection of Hg2+, based upon organic fluorophores7 or chromophores,8 semiconductor nanocrystals,9 cyclic voltammetry,10 polymeric materials,11 proteins,12 and microcantilevers,13 have been developed. Colorimetric methods, in particular, are extremely attractive because they can be easily read out with the naked eye, in some cases at the point of use. Although there are now several chromophoric colorimetric sensors for Hg2+,8 all of them are either limited with respect to sensitivity (current limit of detection≈1 μM) and selectivity, kinetically unstable, or incompatible with aqueous environments. Recently, DNA-functionalized gold nanoparticles (DNA–Au NPs) have been used in a variety of forms for the detection of proteins,14, 15 oligonucleotides,15–21 certain metal ions,22 and other small molecules.23, 24 DNA–Au NPs have high extinction coefficients (3–5 orders of magnitude higher than those of organic dye molecules)25 and unique distance-dependent optical properties that can be chemically programmed through the use of specific DNA interconnects, which allows one, in certain cases,16–20 to detect targets of interest through colorimetric means. Moreover, these structures, when hybridized to complementary particles, exhibit extremely sharp melting transitions, which have been used to enhance the selectivity of detection systems based upon them.16, 18, 20, 26 By using such an approach, one can typically detect nucleic acid targets in the low nanomolar to high picomolar target concentration range in colorimetric format. The ability to use such particles to detect Hg2+ in the nanomolar concentration range in colorimetric format would be a significant advance, especially when one considers that commercial systems for detecting Hg2+ rely on cumbersome inductively coupled plasma approaches that are not suitable for point-of-use applications. Herein, we present a highly selective and sensitive colorimetric detection method for Hg2+ that relies on thymidine–Hg2+–thymidine coordination chemistry27 and complementary DNA–Au NPs with deliberately designed T–T mismatches. When two complementary DNA–Au NPs are combined, they form DNA-linked aggregates that can dissociate reversibly with a concomitant purple-to-red color change.24, 28 For our novel colorimetric Hg2+ assay, however, we prepared two types of Au NPs (designated as probe A and probe B, see the Supporting Information), each functionalized with different thiolated-DNA sequences (probe A: 5′ HS-C10-A10-T-A10 3′, probe B: 5′ HS-C10-T10-T-T10 3′), which are complementary except for a single thymidine–thymidine mismatch (shown in bold; Scheme 1). Importantly, these particles also form stable aggregates and exhibit the characteristic sharp melting transitions (full width at half maximum<1 °C) associated with aggregates formed from perfectly complementary particles, but with a lower melting temperature Tm.17, 18 Since it is known that Hg2+ will coordinate selectively to the bases that make up a T–T mismatch, we hypothesized that Hg2+ would selectively bind to the T–T sites in our aggregates formed from mismatched strands and raise the Tm of the resulting structures.27 The analogous interaction with particle-free DNA leads to significant increases in Tm (ΔTm≈10 °C). Colorimetric detection of mercuric ion (Hg2+) using DNA–Au NPs. The assay begins by adding an aliquot of an aqueous solution of Hg2+ at a designated concentration to a solution of the DNA–Au NP aggregates formed from probes A and B (1.5 nM each) at room temperature (see the Supporting Information). The solution is then heated at a rate of 1 °C min−1 while its extinction is monitored at 525 nm, where the Au NP probes exhibit the maximum intensity in the visible region of the spectrum. The Tm is obtained at the maximum of the first derivative of the melting transition. Without Hg2+, the aggregates melt with a dramatic purple-to-red color change at about 46 °C. In the presence of Hg2+, however, the aggregates melt at temperatures higher than 46 °C because of the strong coordination of Hg2+ to the two thymidines that make up the T–T mismatch, thereby stabilizing the duplex DNA strands containing the T–T single base mismatches. To evaluate the sensitivity of the assay, different concentrations of Hg2+ from one stock solution were tested. When an Hg2+ sample was mixed with the Au NP probe aggregate solution, there was no noticeable change under the reaction conditions described above. Once heated, however, the aggregates melt with a significant purple-to-red color change at a specific temperature (Figure 1 a), which is linearly related to the concentration of Hg2+ over the entire concentration range studied (Figure 1 b). The present limit of detection for this system is approximately 100 nM (=20 ppb) Hg2+ (Figure 1 a), which, to the best of our knowledge, is the lowest ever reported for a colorimetric Hg2+ sensing system. Each increase in concentration of 1 μM results in an increase in Tm by about 5 °C, thus providing an easy way of determining Hg2+ concentration. a) Normalized melting curves of aggregates (probes A and B) with different concentrations of Hg2+. b) Graph of the Tm for the aggregates as a function of Hg2+ concentration. Three components of the assay contribute to its high sensitivity, selectivity, and quantitative capabilities: 1) the oligonucleotides, 2) the Au NPs, and 3) the oligonucleotide–nanoparticle conjugate. From the standpoint of the oligonucleotides, the chelating ability of the thymidines that form the mismatch in the oligonucleotide duplex is extremely selective for Hg2+. It is known that two thymidine residues, when geometrically preorganized in a DNA duplex, can behave as a chelate and form a tightly bound complex with Hg2+.29 From the standpoint of Au NPs, the high extinction coefficients of Au NPs (ca. 108 cm−1 m−1 for 15-nm Au NPs) can act as an amplifier for the perturbation of the Tm upon binding Hg2+, thus allowing detection limits in the ppb range. Conventional chromogenic chemosensors have relatively low extinction coefficients (typically around 105 cm−1 m−1), which limit their sensitivity at best to the micromolar concentration range. Finally, the sharp, highly cooperative melting properties of aggregates made from oligonucleotide–Au NP conjugates enable one to distinguish subtle differences in Tm clearly, thus providing a measure of the Hg2+ concentration from 100 nM to the low micromolar range.16–18 The selectivity of this system for Hg2+ was evaluated by testing the response of the assay to other environmentally relevant metal ions, including Mg2+, Pb2+, Cd2+, Co2+, Zn2+, Fe2+, Ni2+, Fe3+, Mn2+, Ca2+, Ba2+, Li+, K+, Cr3+, and Cu2+ (Figure 2 a and 2b) at a concentration of 1 μM. Only the Hg2+ sample shows a significantly higher Tm (ΔTm≈5 °C) relative to that of the blank. Indeed, at 47 °C, only the aggregate solution containing Hg2+ is purple, whereas all others have turned bright red. Pb2+ is the only other metal ion that influences the Tm of the aggregate, but only by a negligible amount (ΔTm≈0.8 °C). Importantly, this selectivity can be visualized with the naked eye (Figure 2 c). a) Normalized melting curves of the aggregates (probes A and B) in the presence of metal ions (each at 1 μM). b) Graph showing the difference between the Tm of the aggregates of the blank and that of the aggregates with different metal ions: 1: blank; 2: Hg2+; 3: Li+; 4: Cd2+; 5: Ca2+; 6: Ba2+; 7: Mn2+; 8: Mg2+; 9: Zn2+; 10: Ni2+; 11: Fe2+; 12: Co2+; 13: Fe3+; 14: K+; 15: Cr3+; 16: Pb2+; 17: Cu2+. c) Color change of the aggregates (probes A and B, each at 1.5 nM) in the presence of various representative metal ions (each at 1 μM) upon heating from room temperature (RT) to 47 °C. The colorimetric results for Cu2+ are not shown, as the data were taken after initial submission of the manuscript; see the Supporting Information. Because of the thiophilic nature of Hg2+, we considered the possibility that it could be removing the thiolated oligonucleotides from the surface of the gold particle, which could result in nonuniformity of the assay and a potential loss of sensitivity and accuracy. To determine if this was occurring, we utilized fluorophore-labeled oligonucleotides to evaluate the number of DNA strands per particle at various Hg2+ concentrations (0.5, 1, and 2 μM over 8 h; oligonucleotide sequence: 5′ HS-C10-A10-T-A10-(6-FAM) 3′; see the Supporting Information). Significantly, Hg2+ shows no evidence of fluorophore quenching, whereas the gold particle is an excellent quencher of fluorescence. Therefore, if the fluorophore-labeled oligonucleotides are removed from the particles they can be easily identified and quantified by fluorescence spectroscopy. The coverage of DNA at the start of the reaction was determined to be approximately 70 strands per particle by using literature methods.28, 30 The mercuric ion, regardless of the concentrations studied, has very little effect on the surface coverage of the DNA (Table 1). Even at elevated temperature (50 °C), there is less than 10 % loss of DNA from the surface of the particle even after prolonged heating (8 h) (Table 1),30 which suggests that the particle probes will be stable over any reasonable assay conditions. Temp. Portion Mercuric Ion Concentration 0.5 μM 1 μM 2 μM RT in the supernatant 2.1±1.0 1.6±1.3 1.8±1.2 on the particles 68.0±0.9 68.9±1.8 68.1±2.0 50 °C in the supernatant 5.9±0.2 6.0±0.8 6.1±1.3 on the particles 64.7±1.1 64.1±0.8 65.0±2.2 In conclusion, we have developed a colorimetric method to detect Hg2+ using DNA–Au NPs in aqueous media with very high selectivity and sensitivity. This method is enzyme-free and does not require specialized equipment other than a temperature control unit. The concentration of Hg2+ can be determined by the change of the solution color at a given temperature or the melting temperature (Tm) of the DNA–Au NP aggregates. Unlike most of the chemosensors for Hg2+ which have been evaluated in organic media or organic–aqueous mixtures owing to their low water-solubility, the high water solubility of the oligonucleotide-modified gold nanoparticle probes allow this assay to be performed in aqueous media without the need for organic cosolvents. Significantly, this method can in principle be used to detect other metal ions by substituting the thymidine in our study with synthetic artificial bases that selectively bind other metal ions.31 Supporting information for this article is available on the WWW under http://www.wiley-vch.de/contents/jc_2002/2007/z700269_s.pdf or from the author. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.