ARTICLE pubs.acs.org/est
Bead-Based Competitive Fluorescence Immunoassay for Sensitive and Rapid Diagnosis of Cyanotoxin Risk in Drinking Water Hye-Weon Yu,† Am Jang,‡ Lan Hee Kim,† Sung-Jo Kim,† and In S. Kim*,† †
School of Environmental Science and Engineering, Gwangju Institute of Science and Technology (GIST), 261 Cheomdan-gwagiro, Buk-gu, Gwangju 500-712, Republic of Korea ‡ School of Civil and Environmental Engineering, SungKyunKwan University, Suwon 440-746, Republic of Korea
bS Supporting Information ABSTRACT: Due to the increased occurrence of cyanobacterial blooms and their toxins in drinking water sources, effective management based on a sensitive and rapid analytical method is in high demand for security of safe water sources and environmental human health. Here, a competitive fluorescence immunoassay of microcystin-LR (MCYST-LR) is developed in an attempt to improve the sensitivity, analysis time, and ease-ofmanipulation of analysis. To serve this aim, a bead-based suspension assay was introduced based on two major sensing elements: an antibody-conjugated quantum dot (QD) detection probe and an antigen-immobilized magnetic bead (MB) competitor. The assay was composed of three steps: the competitive immunological reaction of QD detection probes against analytes and MB competitors, magnetic separation and washing, and the optical signal generation of QDs. The fluorescence intensity was found to be inversely proportional to the MCYSTLR concentration. Under optimized conditions, the proposed assay performed well for the identification and quantitative analysis of MCYST-LR (within 30 min in the range of 0.4225 μg/L, with a limit of detection of 0.03 μg/L). It is thus expected that this enhanced assay can contribute both to the sensitive and rapid diagnosis of cyanotoxin risk in drinking water and effective management procedures.
’ INTRODUCTION Surface water sources such as streams, rivers, lakes, and oceans are the most important drinking water resources. However, internationally, these water reservoirs have become contaminated and now suffer from eutrophic conditions due to excess nutrient loading from anthropogenic origins. For example, the accelerated eutrophication of surface water and climate change may lead to an increasing frequency of cyanobacterial blooms;1 globally, cyanobacterial blooms are regarded as one of the most serious public health issues due to the liberation of their toxins and the deterioration of water quality in drinking water sources.2 Among cyanotoxins, the most prevalent is a hepatotoxin of microcystin (MCYST). Cases of adverse health effects from MCYST have been reported in many countries, with some reports of liver cancer and even death3 due to its severe poisoning of animals and humans.4,5 To date, approximately 80 variants of MCYST heptapeptides have been discovered6—the most common and toxic being MCYST-LR, in which leucine (L) and arginine (R) are combined with a constant ring composed of five amino acids (Figure 2b). Since MCYST-LR can be considered a surrogate for toxicity, equivalent to MCYST analogs in actual and acute toxicity, the World Health Organization (WHO) has proposed a provisional upper limit for total MCYST-LR (intracellular and extracellular toxins) of 1 μg/L in drinking water.7 As such, it poses a major challenge in the production of safe r 2011 American Chemical Society
drinking water from surface waters containing cyanobacteria with their toxin. MCYST-LR may occur within cells (cell-bound or intracellular) or be released into water (dissolved or extracellular) due to natural cell lysis or via active release, and their properties strongly influence the removal efficiency in drinking water treatment plants (Table S1 in the Supporting Information). The approach to eliminate the microbial cell and its toxin as an adsorbate by activated carbon suffers from the decrease of adsorption capacity caused by operational problems such as requirement of long reaction time, competition with natural organic matter (NOM) for adsorption sites, and pore blockage.8 Because MCYST-LR is extremely stable across a wide range of pH and temperature due to its cyclic nature, cyanotoxins remain after oxidative degradation with strong oxidants and UV exposure within the range of normal disinfection practices.9 In addition, the high doses of oxidants required to completely decompose the compound can lead to the formation of disinfection byproducts such as trihalomethanes or bromate.9 In addition, in membrane-based treatment processes, the exclusion ability of UF and NF having relatively large pore size is insufficient to successfully Received: April 19, 2011 Accepted: July 26, 2011 Revised: June 23, 2011 Published: August 18, 2011 7804
dx.doi.org/10.1021/es201333f | Environ. Sci. Technol. 2011, 45, 7804–7811
Environmental Science & Technology
ARTICLE
Figure 1. Conceptual scheme for competitive fluorescence immunoassay.
separate extracellular MCYST-LR from intake water due to its low molecular weight (approximately 1 kDa). And the shear force resulting from the hydraulic pressure-driven flow may also induce cell damage and an increased amount of dissolved toxins. Therefore, the most effective way to remove MCYST-LR in the drinking water supply is to remove intact cyanobacterial cells that retain the majority of metabolites inside the cells, without damaging them.10 Then, to successfully control the unavoidable soluble toxins caused by cell damage, a sensitive and rapid analysis of MCYST-LR should take priority, by monitoring its quantity over a period that includes its intake and finishes with the final production of drinking water. Because common analytical techniques for determining MCYST-LR concentration (Table S2) suffer from drawbacks with respect to their narrow dynamic range and long analysis time, and though they satisfy the limit of detection established by WHO guidelines, they are not suitable for practical applications in routine environmental monitoring conditions. In this study, we develop a more sensitive and rapid analytical tool for detecting MCYST-LR using a competitive fluorescence immunoassay. In an attempt to improve upon existing immunoassay techniques such as the enzyme-linked immunosorbent assay (ELISA) regarding their sensitivity, overall analysis time, and simplified manipulation, a bead-based suspension assay was introduced by combining fluorescence nanocrystals, quantum dots (QDs), and magnetic beads (MBs). Compared to the use of enzyme-catalyzed signal generation to convert an end-product chromophore, the optical response of QDs can be easily and directly measured as photoluminescence (i.e., fluorescence). Because QDs have superior optical advantages such as high quantum yield, high molar extinction coefficients, and high photostability,11 they can considerably enhance the assay sensitivity with an optically amplified signal.12 Hence, antibody-conjugated QDs were utilized as a detection probe for the immunological recognition of MCYST-LR,13 as well as for optical transduction. Furthermore, antigen-coated MBs were involved in a competitive reaction with an analyte, which facilitated the separation step. Based on this well-validated assay performance, the competitive fluorescence immunoassay enabled the analysis of intracellular and extracellular MCYST-LRs produced from cyanobacteria, as a means of elucidating the toxin production kinetics.
’ MATERIALS AND METHODS Preparation of QD Detection Probe. Monoclonal antibodies (mAbs) against MCYST-LR (MC10E7, Enzo Life Sciences) were coated on the surface of QDs (Qdot525 antibody conjugation kit, Invitrogen) in accordance with the manufacturer’s protocol, with minor modifications as described in the Supporting Information-1 (SI-1). In our previous study, the binding capacity of the synthesized QD detection probe was estimated to be 10 molecules of MCYST-LR per one probe.14 Synthesis of MB Competitor. An MB competitor was prepared by tethering MCYST-LR to magnetic beads through a serial twostep reaction: the biotinylation of MCYST-LR and the conjugation of biotinylated toxin on a streptavidin-coated MB surface. Next, 0.2 μmol of MCYST-LR (Alexis) was mixed with 0.3 μmol of N-hydroxysuccinimide (NHS) and 0.3 μmol of dicyclohexylcarbodiimide (DCC) dissolved in anhydrous dimethylsulfoxide (DMSO)/acetonitrile (25:75% v/v) in a desiccator overnight. The activated NHS esters of the carboxylic acids were then reacted with the primary amine groups of 0.4 μmol EZ-link Amine-PEG3Biotin (Thermo Scientific) for 21 h, and then incubated with triethylamine (0.3 μmol) for 3 h to quench the remaining unreacted amine-reactive groups. The above reagent proportion was selected in order to minimize protein polymerization, by using a large molar excess of amine-PEG3-biotin and a limiting amount of DCC. In addition, 500 μL of streptavidin-coated superparamagnetic beads (MagnaBind streptavidin beads, Pierce), corresponding to the biotin binding capacity of 4 nmol (50-fold molar excess over biotinylated MCYST-LR), was incubated with the previously prepared biotin-modified peptides suspension for 24 h under constant mixing. The unbound materials were then removed from the synthesized MB competitor by several washings with PBS buffer using a magnetic particle concentrator (Invitrogen). The final complex was then resuspended and kept in 0.02% sodium azide/PBS as a storage buffer at 4 °C prior to use. Optimization of Competitive Immunological Reaction. Experimental factors involving competitive immunological reactions were optimized to improve assay performance pertaining to sensitivity and analysis time. First, the binding capacity of the MB competitor was estimated to determine the maximum ability for capturing the QD detection probe on its MCYST-LR-immobilized surface. In 7805
dx.doi.org/10.1021/es201333f |Environ. Sci. Technol. 2011, 45, 7804–7811
Environmental Science & Technology
ARTICLE
Figure 2. (a) Conceptual strategy for the synthesis of MB competitor using the biotinylation of MCYST-LR and subsequent immobilization on a streptavidin-coated MB, and (b) molecular structure of MCYST-LR.
addition, reagent concentrations of the QD detection probe (2.0 108 to 2.5 109 M) and MB competitor (8.3 1014 to 5.2 1015 M), incubation time (1090 min), and incubation temperature (2040 °C) were optimized in order to investigate the best experimental conditions for competitive immunological recognition and further generation of the QD fluorescence signal. To determine the optimum experimental conditions, a sensitivity index (B/Bo) was introduced, which is described as the ratio between the optical signal (fluorescence) obtained in the presence of the analyte (B) and the signal in the absence of analyte (Bo).15 The optimal factors that provide the lowest sensitivity index along with the steepest slope were subsequently selected for further analysis. Competitive Fluorescence Immunoassay. Standard MCYSTLRs were prepared via serial dilution with 100% methanol, in a range of 0.001100 μg/L. After addition to 50 μL of 4.2 1014 M MB competitor, they were then incubated with 50 μL of 5.0 109 M QD detection probe for 30 min at room temperature. Magnetic separation and washing were then thoroughly carried out with PBS to remove unbound materials. After excitation (λex = 450 nm), the final stage of the assay involved the fluorescence measurement of QDs at 533 nm, using an F-2500 fluorescence spectrophotometer (Hitachi).
’ RESULTS AND DISCUSSION Competitive Fluorescence Immunoassay for MCYST-LR Detection. As MCYST-LR is a relatively small oligopeptide
accounting for only 1 kD (C49H74N10O12), it is difficult to apply a sandwich immunoassay due to the steric hindrance of available binding sites on the surface of a target antigen using double antibodies. Hence, a competitive immunoassay was designed to identify small molecules using a single antibody.15 In this study, two major sensing elements (an Ab-coated QD detection probe and an MCYST-LR-immobilized MB competitor) were introduced to compete with MCYST-LR as analyte. The purpose of the Ab-coated QD detection probe was to provide (1) immunological binding to specifically recognize the arginine (R) moiety at position 4 of MCYST-LR, and (2) optical signal generation. The MCYST-LR-immobilized MB competitor also has two functions: (1) competitive binding with analytes for the limited binding sites of the QD detection probe, and (2) separation of nonbound materials as a solid phase. First, the MB competitors were added to the MCYST-LR sample (Figure 1, step 1). QD detection probes were then allowed to react with the previous solution, which included both the analyte and MB competitor (Figure 1, step 2). During incubation, the QD detection probes were relatively partitioned between the free MCYST-LRs (analyte) in solution and antigens bound to the MB competitor; both reactions occurred simultaneously based on the equilibrium constant. In competitive assays, the equilibrium constant is the major limiting factor due to fractionation between the antigen and antibody due to the law of mass action. Once the equilibrium state of this competitive antigenantibody reaction was reached, the suspension was 7806
dx.doi.org/10.1021/es201333f |Environ. Sci. Technol. 2011, 45, 7804–7811
Environmental Science & Technology
ARTICLE
placed under an activated external magnetic field, and the QD detection probe-tethered MB competitors were accumulated at one spot using a magnet. The MB conjugates were then separated from the liquid phase, and included unbound components such as nonspecific biomolecules, excess analyte, and QD detection probes or the QD detection probe-conjugated analyte (Figure 1, step 3). After supernatant removal, the MB retentate was transferred to an optical transduction unit to measure the QD fluorescence. The final stage of the assay involved a quantitative analysis of MCYST-LR from the green fluorescence intensity of the QDs at the specific wavelength in response to the photon excitation energy (Figure 1, step 4 and Figure S1). As a result, we found that the optical signal produced from the QDs, which was the competitive proportion to couple with the MB
Figure 3. Binding capacity of MB competitor against QD detection probes. The data marked with the asterisk indicates the theoretical amount of streptavidin molecules coated on the MB surface, which corresponds to the equivalent amount of biotinylated MCYST-LR and, consequently the expected binding capacity. The value in the dashed box is estimated to be the actual binding capacity for QD detection probes.
competitor, was inversely proportional to the concentration of MCYST-LR analyte in the sample. Characterization and Optimization of Competitive Fluorescence Immunoassay. Once the assay format and two major sensing elements were set up, the goal of immunoassay optimization is to improve the sensitivity and to reduce the analysis time, which are mainly governed by reagent characteristics (the binding capacity or affinity of two sensing components: the QD detection probe14 and MB competitor, and by assay conditions (reagent concentration, incubation time, and incubation temperature). Binding Capacity of MB Competitor. To fabricate the MB competitor, the biotinylation method was used for tagging MCYST-LRs, which were subsequently immobilized to the streptavidin molecules that coated the MB surface in the following conjugation process. The hydrophobic heptapeptide MCYST-LR was first reconstituted with biotin molecules by combining the biotinylation reagent with the activating reagent (NHS) and the coupling reagent (DCC) in an organic base solvent. The biotin compound used in this study was amine-PEG3-biotin, which contains a terminal primary amine and a polyethylene glycol (PEG) spacer arm. The amine group of this reagent can be reacted with two reactive carboxylic groups on the C-terminal, D-Glu (glutamic acid), and D-MeAsp (aspartic acid) moieties of MCYST-LR (Figure 2b).16 The intermediary PEG spacer arm also provides a long and flexible connection to minimize the steric hindrance and loss of biological active function involved with antigenantibody binding reaction, to further promote QD detection probe labeling (Figure 2a).17 Determining the binding capacity of the MB competitor is crucial for evaluating the immunological affinity against an antibody and the comprehensive titration of the QD detection probe. Here, the binding affinity of the MB competitor was thus estimated by measuring the QD fluorescence intensity (%) when excess and constant QD detection probes were reacted to seven
Table 1. Experimental Optimization of Reagent Concentration of Two Sensing Components Using the Checkerboard Titration Method QD/Ab detection probe (nM) MB/mcyst competitor (fM)
20.0
10.0
5.0
2.5
1. total binding/nonspecific bindinga 83.0
c
1.30 (7.31)
1.28 (9.39)
1.27 (5.76)
1.27 (4.14)
41.5
1.73d (1.06)
1.71 (6.85)
1.70 (3.19)
1.68 (7.39)
20.8 10.4
1.79 (6.53) 1.73 (8.15)
1.75 (0.95) 1.68 (5.69)
1.63 (4.42) 1.57 (1.54)
1.59 (3.71) 1.57 (5.03)
5.2
1.65 (1.64)
1.65 (4.93)
1.63 (3.78)
1.61 (4.41)
b
2. sensitivity index (B/Bo) 83.0
0.98 (2.70)
0.92e (4.59)
0.99 (2.27)
0.96 (2.22)
41.5
1.01 (5.24)
0.93f (3.54)
0.89 (3.17)
1.01 (4.16)
20.8
0.98 (5.10)
0.97 (5.93)
0.99 (6.06)
0.94 (8.84)
10.4
1.02 (6.06)
0.99 (4.45)
1.03 (3.04)
0.97 (2.77)
5.2
1.00 (1.52)
1.00 (0.59)
0.99 (2.01)
1.01 (2.47)
a
Total binding was obtained in the absence of analyte at varying combinations of QD detection probe and MB competitor, whereas nonspecific binding was estimated only in the absence of QD detection probe. b Sensitivity index was calculated from the QD fluorescence intensity with 2.5 ppb MCYST-LR (B) and in the absence of target antigen (Bo). c The parenthesized values indicate %CV (coefficient of variation) described as the ratio of standard deviation to average with three replicate experiments. d The underlined values show a reasonable S/N (>1.7) for the total binding reaction between two components. e The values in bold indicate an acceptable sensitivity index that satisfies the determined criteria (
