Article pubs.acs.org/ac
Trace-Level Mercury Ion (Hg2+) Analysis in Aqueous Sample Based on Solid-Phase Extraction Followed by Microfluidic Immunoassay Yasumoto Date,*,†,‡ Arata Aota,† Shingo Terakado,† Kazuhiro Sasaki,† Norio Matsumoto,†,‡ Yoshitomo Watanabe,†,‡ Tomokazu Matsue,‡,§ and Naoya Ohmura† †
Environmental Science Research Laboratory, Central Research Institute of Electric Power Industry, 1646 Abiko, Abiko City, Chiba 270-1194, Japan ‡ Graduate School of Environmental Studies, Tohoku University, 6-6-11, Aramaki, Aoba, Sendai 980-8579, Japan § The World Premier International Research Center Initiative, Advanced Institute for Material Research (WPI-AIMR), Tohoku University, 2-1-1 Katahira Aoba, Sendai, 980-8577 Japan S Supporting Information *
ABSTRACT: Mercury is considered the most important heavymetal pollutant, because of the likelihood of bioaccumulation and toxicity. Monitoring widespread ionic mercury (Hg2+) contamination requires high-throughput and cost-effective methods to screen large numbers of environmental samples. In this study, we developed a simple and sensitive analysis for Hg2+ in environmental aqueous samples by combining a microfluidic immunoassay and solid-phase extraction (SPE). Using a microfluidic platform, an ultrasensitive Hg2+ immunoassay, which yields results within only 10 min and with a lower detection limit (LOD) of 0.13 μg/L, was developed. To allow application of the developed immunoassay to actual environmental aqueous samples, we developed an ion-exchange resin (IER)-based SPE for selective Hg2+ extraction from an ion mixture. When using optimized SPE conditions, followed by the microfluidic immunoassay, the LOD of the assay was 0.83 μg/L, which satisfied the guideline values for drinking water suggested by the United States Environmental Protection Agency (USEPA) (2 μg/L; total mercury), and the World Health Organisation (WHO) (6 μg/L; inorganic mercury). Actual water samples, including tap water, mineral water, and river water, which had been spiked with trace levels of Hg2+, were well-analyzed by SPE, followed by microfluidic Hg2+ immunoassay, and the results agreed with those obtained from reduction vaporizing−atomic adsorption spectroscopy.
A
inductively coupled plasma−mass spectroscopy (ICP-MS).10 Although these instrumental analyses are sensitive and accurate, they are also time-consuming, expensive, and require complex pretreatment and laboratory equipment. To overcome these limitations, simplified mercury analysis methods have also been investigated. Recently, DNAzyme-,11,12 oligonucleotide-,13 and functional nanoparticle-based14,15 measurements have been reported; however, these methods are not practical, because chemically closely related metals interfere with the quantitative measurements. By contrast, immunoassays offer the potential for simple, fast, and cost-effective measurements, but they require the development of suitable antibodies and protocols. To date, a large number of antibodies specific for heavy-metal species have already been identified.16−21 A mercury-specific monoclonal antibody has been reported using mercury-organic compounds (EDTA, glutathione, and 6-mercaptonicotinic acid) as hapten.22−24 Our group has also identified antibodies specific for Hg2+-EDTA.
mong various environmental pollutants, mercury is considered the most important;1 because of its bioaccumulation and toxicity, this heavy metal is extremely dangerous for all biological organisms, and for humans in particular.2,3 Mercury is present in surface and groundwater at trace concentration levels as ionic form (Hg2+), although local mineral deposits and contaminants may produce higher levels in groundwater. Methylation of the mercury ion to form highly toxic organic mercury (such as CH3Hg+) occurs in fresh water as well as seawater;4 this methylated mercury is readily bioaccumulated in the ecological chain.5,6 The World Health Organisation (WHO) suggested 6 μg/L of inorganic mercury as the guideline tolerable value for mercury in drinking water, while the United States Environmental Protection Agency (USEPA) guideline value is 2 μg/L. Mercury contamination is widespread in the environment, and monitoring of mercury in the environment requires a large number of measurements at regular intervals. Therefore, monitoring of mercury in the environment requires highthroughput and cost-effective methods to enable the screening of very large numbers of samples. The methods most frequently used in environmental analysis of mercury are atomic absorption spectroscopy (AAS), 7,8 atomic fluorescence, 9 and © 2012 American Chemical Society
Received: November 5, 2012 Accepted: November 27, 2012 Published: November 27, 2012 434
dx.doi.org/10.1021/ac3032146 | Anal. Chem. 2013, 85, 434−440
Analytical Chemistry
Article
combined with microfluidic immunoassay and compared to conventional instrumental analysis.
Although several immunoassays based on ELISA and immunochromatography of mercury and application of these approaches to environmental samples has been reported,25−27 almost none of these investigations used pretreatment to increase the specificity and efficacy of the antibody. For actual environmental research, unknown coexisting metals, which may cross-react with the antibody, may be present at high levels, and unforeseeable organic and inorganic compound may disturb the accuracy of the assay, even though the pH of the solution is adjusted. The most commonly used preparation method for aqueous samples are liquid−liquid extraction and solid-phase extraction (SPE). Especially, SPE does not require solvents, and it allows for cleanup and enrichment of the sample in a single step. Furthermore, it is possible to exchange the sample solution with any desired solution that may be beneficial for subsequent measurements, such as a neutral buffer suitable for immunoassays. Given the examples of previous investigations using immunoassays for the detection of pesticides in environmental samples, SPE followed by immunoassay appears to be a cost-effective and useful approach for extracting the analyte and removing impurities, including many cross-reactive impurities.28 For selective extraction of mercury before instrumental measurement, the use of ion-exchange resin (IER) and chelating resin7,29−33 have also been reported. IER-based SPE was less selective than a chelating resin, but IER is more cost-effective and is relatively easy to elute with a weak elution solvent. Microfluidic devices possess many of the features that make bioassays advantageous, including high throughput, short analysis time, and the ability to operate with small sample volumes and high sensitivity, because of their microfluidic properties, such as reduction of the incubation and mixing times.34,35 Therefore, microfluidic immunoassays offer the possibility of miniaturization, integration, and automation, which facilitate low-cost and simple measurements, and are therefore a promising technique for use in the fields of environmental and food safety surveillance, as well as clinical diagnostics.36−39 Several different detection systems for microfluidic immunoassays such as fluorescent detection, electrochemical detection, and surface plasmon resonance spectroscopy have developed, and each has its advance points;40−42 however, each of these methods requires expensive and bulky instrumentation. Absorbance measurements are a potentially useful detection method to develop a compact optical system, low-cost apparatus for a microfluidic immunoassay if sufficient material can be obtained. Recently, we developed a microfluidic immunoassay based on the accumulation of AuNP-labeled antibody and simple absorbance measurement, which enabled an antibody dissociation constant (Kd)-limited immunoassay within a few minutes.43 In this study, we developed a practical method for the analysis of trace levels of Hg2+ in environmental aqueous samples by SPE-based pretreatment and a rapid microfluidic-based immunoassay. First, the antibody recognizing Hg2+-EDTA was characterized and applied to the microfluidic immunoassay, which enables antibody Kd-limited detection. In order to achieve accurate Hg2+ analysis in environmental samples, selective Hg2+ extraction from a mockup solution containing coexisting metal ions (10 environmental metals: aluminum [Al3+], cadmium [Cd2+], chromium [Cr3+], copper [Cu2+], iron [Fe3+], manganese [Mn2+], magnesium [Mg2+], lead [Pb2+], zinc [Zn2+], and calcium [Ca2+]) was performed by SPE using ion-exchange resin. Finally, tap water, mineral water, actual river water, and highly contaminated mockup river water, were analyzed by SPE
■
EXPERIMENTAL SECTION Reagents. Strong base cation-exchange resin (Dowex 1 × 2, 100−200 and 200−400, Cl form) was purchased from Muromachi Technos; gold chloride acid (HAuCl4), ionic mercury standard solution (HgCl2:100 mg/L), and multielement standard solution W-6 (Al3+, Ca2+, Mg2+, Cd2+, Cr3+, Cu2+, Fe3+, Mn2+, Pb2+, and Zn2+, each 100 mg/L) was obtained from Wako Pure Chemicals, and 1-(4-isothiocyananobenzyl) ethylenediamine-N,N,N′,N′-tetraacetic acid (ITCB-EDTA) was purchased from Dojindo Laboratories. River water samples (JSAC 0301-3 and JSAC 0302-3) were purchased from the Japan Society for Analytical Chemistry. All other chemicals were of analytical grade and were used as received. Aqueous solutions were prepared using distilled, deionized water obtained from a Milli-Q filtration system (Millipore, Bedford, MA). Antibodies. The monoclonal antibody NX2C3, which recognizes Hg2+-EDTA complexes, was prepared from mouse hybridomas, as previously described.20 Goat antimouse IgG (H+L) was purchased from Jackson Immunoresearch, Inc. To prepare the AuNP-labeled secondary antibodies, 0.2 mL of antibody (1.5 mg/mL) was added to 200 mL of the monodispersed AuNP (40 nm in size) solution, which adjusted to pH 9.15 (above the isoelectric point (IEP) of the antibody), and thereafter, the solution was mixed gently. After the addition of bovine serum albumin (BSA) solution to a concentration of 1%, to block the AuNP surface, the conjugate was centrifuged at 9000 rpm (18000g) for 25 min, and then, the sediment (15 mL from 200 mL original solution) was collected. Preparation of the Antigen Immobilized Beads. The beads with immobilized antigen was prepared as described previously.43 In brief, 0.4 g of PMMA beads (100 μm, Gantsu-kasei Corp., Japan) were suspended in 1 mL of PBS containing 10 mg of BSA, and were then gently mixed for 2 h. The overlying solution was then removed, and the beads were washed 4 times with 100 mM boric acid (pH 9.8), after which the beads were physically adsorbed with BSA. The beads were then resuspended in 0.8 mL of boric acid and were then mixed with 165 μL of 1 mg/mL ITCB-EDTA solution for 2 h. The beads were then washed 5 times with 50 mM MES buffer (pH 6.5), followed by the addition of 195 μL of 20 mM HgCl2 solution. After incubation for 30 min and washing five times with buffer, Hg2+-chelate immobilized beads were stored at 4 °C before measurement. Microfluidic Device. The microfluidic device for flowbased immunoassay was fabricated as previously described.43 The microfluidic device consisted of a main channel for introducing the sample solution and washing buffer, and a subchannel for manipulating the beads. The detection area, which aligned with the optical source (LED) and photodiode, was defined by damlike structures with barriers 100 μm wide at intervals of 50 μm, as shown in Figure 1A. The damlike structures trap a limited amount of beads within the detection area (w × l × d: 1000 μm × 1000 μm × 225 μm) and facilitate the interaction between the antibody and the beads under uniform flow conditions. Microfluidic Immunoassay. The measurement scheme of microfluidic immunoassay was as follows. First, antigenimmobilized beads were packed into the defined detection area. Second, the sample solution, containing anti-Hg2+-EDTA antibody and EDTA, was introduced via the main channel and 435
dx.doi.org/10.1021/ac3032146 | Anal. Chem. 2013, 85, 434−440
Analytical Chemistry
Article
The limit of detection (LOD) was calculated based on the 3-sigma method. Solid-Phase Extraction. The SPE column consisted of a 5-mL plastic syringe (TERUMO), packed with layers of ionexchange resin sandwiched between two circular filter papers (5B, diameter = 13 mm, Advantech). In order to absorb the Hg2+ on and exclude the coexisting metal ion species that may interfere with the immunoassay, a strong base anion-exchange resin, Dowex 1 × 2 (chloride form), was used. The suspension of ion-exchange resin (resin:water = 1:1 [vol/vol]) was spread on a filter paper at the bottom of the column, and was then covered with the second filter. The prepared pretreatment column was used immediately to avoid drying and contamination. Elution of the column required 10−12 min after the addition of 5 mL of solution, regardless of the amount of packed resin. Therefore, the flow rate of the solution through the column was calculated as 400−500 μL/min. Before addition of the sample, the SPE column was equilibrated with 5 mL of HCl acid solution, which was also used to adjust the concentration of the samples. In order to adsorb Hg2+ and reduce the level of adsorption of coexisting ions, a series of samples adjusted with various concentrations of HCl acid were tested. Adsorbed Hg2+ was eluted with several eluent after column washing. Instrumental Analysis. Mercury level was determined using a cold-vaporizing atomic absorption spectrometry (CVAAS) device (Mercury Analyzer HG-200, Hiranuma Sangyo Co., Ltd., Japan). Other metal ions were determined using inductively coupled plasma−atomic emission spectrometry (ICP-AES, Varian, Inc.). The wavelengths (nm) for analysis of each element were as follows: Mn, 257.610; Mg, 279.553; Fe, 238.204; Zn, 213.857; Pb, 220.353; Al, 396.152; Cd, 214.439; Cu, 324.754; Cr, 267.716; Ca, 396.847. Actual Sample Analysis. Tap water was collected from the local water works in Abiko, Japan. Mineral water was obtained from commercial bottled natural spring water (Volvic, KIRIN MC DANONE WATERS Co.,Ltd.). River water reference samples (JSAC0301-3 and JSAC 0302-3) were obtained from Seishin Trading Co. Ltd. (Japan). JSAC0302-3 was river water (JSAC0301-3) spiked with 13 inorganic elements (Pb2+, Cr3+, Cd2+, Se3+, As3+, Cu2+, Fe3+, Mn2+, Zn2+, B3+, Al3+, Ni2+, and Be2+). All water samples were adjusted to the optimized concentration of HCl, and were then extracted by SPE. The collected eluates were then immediately analyzed by microfluidic immunoassay. Hg2+ concentration was determined by both immunological and instrumental approaches, using a standard curve method.
Figure 1. Schematic of the flow-based mercury immunoassay. (A) Schematic depiction of microfluidic device and configuration for absorbance measurement. Detection are in microfluidic device was packed with antigen immobilized beads (inset). (B) Accumulation of Hg2+-EDTA specific antibody on antigen-immobilized beads during flow condition. (C) Reaction between the accumulated primary antibody and AuNP-labeled secondary antibody.
passed rapidly through the detection area packed with the beads (Figure 1B). Under the flow of sample solution, a small portion of the free antibody in the sample was captured and accumulated on the surface of the beads in the detection area. Third, in order to determine the primary antibody, 4 nM of AuNP-labeled secondary antibody was introduced (Figure 1C). Fourth, uncaptured antibody was removed by perfusion of 1 mL of PBS containing 1% of blocking reagent (N101, NOF Corporation). To evaluate the amount of captured antibody, absorbance change at the detection area by the accumulated AuNP was determined. The light passing through the antigenimmobilized beads was recorded as the voltage output from the photodiode before (I0) and after (I) the introduction of the sample. Absorbance (A) was calculated using the equation
■
RESULTS AND DISCUSSION Antibody Characterization. The Kd value for Hg2+ and the other metals (Al3+, Cd2+, Cr2+, Cu2+, Fe3+, Mn2+, Mg2+, Pb2+, Zn2+, and Ca2+) were evaluated using fluorescent-based commercial flow immunoassays (KinExA3200, Sapidyne, Inc.), and then cross-reactivity was calculated as shown in Table 1. The antibody Kd value for Hg2+ was 0.89 μg/L (4.4 nM), which enables detection of nanomolar levels of Hg2+ in the sample. Significantly low cross-reactivity (99.5%), regardless of the HCl concentration. In addition, Hg2+ adsorption gradually decreased with an increase in HCl concentration when 50 μL of IER was used. Therefore, 100 μL of IER was used for the solid phase in the following experiments. To investigate quantitative recovery of the adsorbed Hg2+ with the smallest possible volume of eluent, we examined several elution solvents. First, 40 mM of HCl and 40 mM of HNO3 were selected as acidic eluents. After the 5 mL of sample containing Hg2+ (500 ng) was loaded onto SPE column, the adsorbed Hg2+ was eluted with 25 mL elution solvent. Five milliliters of each fraction was collected and analyzed using CV-AAS. HCl acid was not able to elute Hg2+ at all (
