Signal-On Photoelectrochemical Immunoassay for Aflatoxin B1 Based

Apr 14, 2017 - With the improvement of living conditions and the development of science and technology, people pay more attention to food safety and ...
1 downloads 6 Views 4MB Size
Download PDF
Article pubs.acs.org/ac

Signal-On Photoelectrochemical Immunoassay for Aflatoxin B1 Based on Enzymatic Product-Etching MnO2 Nanosheets for Dissociation of Carbon Dots Youxiu Lin,† Qian Zhou,† Dianping Tang,*,† Reinhard Niessner,‡ and Dietmar Knopp*,‡ †

Key Laboratory of Analysis and Detection for Food Safety (MOE & Fujian Province), Collaborative Innovation Center of Detection Technology for Haixi Food Safety and Products (Fujian Province), State Key Laboratory of Photocatalysis on Energy and Environment, Department of Chemistry, Fuzhou University, Fuzhou 350108, People’s Republic of China ‡ Chair for Analytical Chemistry, Institute of Hydrochemistry, Technische Universität München, Marchioninistrasse 17, München D-81377, Germany ABSTRACT: Aflatoxin B1 (AFB1) monitoring has attracted extensive attention because food safety is a worldwide public health problem. Herein, we design a novel simultaneously visual and photoelectrochemical (PEC) immunosensing system for rapid sensitive detection of AFB1 in foodstuff. The immunoreaction was carried out on anti-AFB1 antibodymodified magnetic beads by using glucose oxidase (GOx)labeled AFB1-bovine serum albumin (AFB1−BSA) conjugates as the tags with a competitive-type immunoassay format, while the visual and PEC evaluation was performed via carbon quantum dots (CQDs)-functionalized MnO2 nanosheets. Accompanying the formation of immunocomplexes, the carried GOx initially oxidized the substrate (glucose) for the generation of H2O2, which reduced/etched MnO2 nanosheets into Mn2+ ions, thereby resulting in the dissociation of CQDs from the electrode. Within the applied potentials, the photocurrent of MnO2CQDs-modified electrode decreased with the increasing H2O2 level in the detection cell. Meanwhile, a visual detection could be performed according to the change in the color of MnO2-CQDs-coated electrode. To elaborate, this system was aggregated into a high-throughput microfluidic device to construct a semiautomatic detection cell. Under optimal conditions, the photocurrent increased with the increasing target AFB1 within a dynamic working range from 0.01 to 20 ng mL−1 with a limit of detection (LOD) of 2.1 pg mL−1 (ppt). The developed immunoassay exhibited good reproducibility and acceptable accuracy. In addition, the method accuracy relative to AFB1 ELISA kit was evaluated for analyzing naturally contaminated or spiked peanut samples, giving the well-matched results between two methods. Although our strategy was focused on the detection of target AFB1, it is easily extended to screen other small molecules or mycotoxins, thereby representing a versatile immunosensing scheme.

W

photoactive material plays a major role in the system. Nowadays, the photoactive material of PEC sensors mainly depends on the photoactive metallic semiconductors (e.g., CdS, TiO2, CdSe, and CdTe) on the basis of their efficient photoelectric conversion elements and unique biological compatibility for the biomolecules.12 However, the potential actual application of metallic semiconductors on the PEC immunoassay is still restricted, which is mainly ascribed to the limitations: (i) inherently high toxicity and poor stability of the commonly used semiconductors in PEC detection such as CdTe and CdSe;13 and (ii) some semiconductors with a wide band gap (e.g., 3.5 eV for SnO2 and 3.2 eV for ZnO) possess strong oxidation characteristics and demand a high-energy excitation optical source (e.g., UV light).14,15 As a consequence, nontoxicity, low cost, and low-energy excitation of photoactive materials are highly desirable for the development of the PEC

ith the improvement of living conditions and the development of science and technology, people pay more attention to food safety and environmental protection.1 Developing a fast and high-selectivity method for the detection of trace analyte is of great interest. The photoelectrochemical (PEC) sensing method is a newly grown and booming detection technique.2,3 During the PEC detection process, light is employed as the excitation source, while the generated photocurrent on the electrode is used as the detection signal.4,5 Because of the combination of photo irradiation with electrochemical detection, PEC sensors possess the strengths of both optical methods and electrochemical sensors.6 First, the high-efficiency energy forms separation of the excitation source and detection signal in the PEC methods, ensuring it owns high sensitivity and low background signals.7 Second, the application of electrochemical signal readout makes the PEC instrument simple, of easy miniaturization, of low power requirements, and low cost.8 Moreover, the PEC sensing system with enhanced photocurrent intensity and less electron−hole recombination is desirable for highly sensitive detection.9−11 Among them, © 2017 American Chemical Society

Received: March 14, 2017 Accepted: April 14, 2017 Published: April 14, 2017 5637

DOI: 10.1021/acs.analchem.7b00942 Anal. Chem. 2017, 89, 5637−5645

Article

Analytical Chemistry

Scheme 1. Schematic Illustration of Photoelectrochemical Immunosensing Platform toward Aflatoxin B1 (AFB1) on Carbon Quantum Dots-Coated MnO2 Nanosheets (MnO2-CQDs) by Coupling an Enzyme Immunoassay Format with a MagnetoControlled Microfluidic Devicea

a

(A) Magnetic competitive immunoassay format on monoclonal anti-AFB1 antibody-conjugated magnetic bead (MB-mAb) using glucose oxidase (GOx)-labeled AFB1-bovine serum albumin (AFB1−BSA) conjugate as the tag and (B) H2O2-responsive dissolution of MnO2 nanosheets from MnO2-CQDs with photocurrent measurement.

intensity on the substrate surface, and maintain the high efficiency of PEC sensors. As is well-known, MnO2 nanostructures are usually reduced to Mn2+ by acidic H2O2 due to pH-/redox-responsive properties.29 Yuan’s group discovered that MnO2 nanosheets can effectively quench the fluorescence of upconversion nanoparticles (UCNPs), while it could be recovered by adding H2O2.30 The recovery in the fluorescence was attributed to the H2O2-mediated redox reaction of reducing MnO2 to Mn2+, resulting in the decomposition of MnO2 nanosheets and the release of the free fluorescent UCNPs. To this end, our motivation of this work is exploring a novel PEC immunosensing strategy focusing on CQDs-functionalized MnO2 nanosheets (MnO2-CQDs) by coupling with H2O2-responsive decomposition of MnO2 nanosheets. Aflatoxins, the highly toxic secondary metabolites produced by a number of different fungi, are present in a wide range of food and feed commodities.31 The major aflatoxins of interest are denoted as B1, B2, G1, and G2; however, aflatoxin B1 (AFB1) is usually predominant and the most hazardous.32 Herein, we report the proof-of-concept of a high-throughput PEC immunosensing system for quantitative monitoring of AFB1 (as a model mycotoxin) by coupling with a semiautomatic microfluidic device (Scheme 1). Glucose oxidase (GOx) is used for the labeling of AFB 1−bovine serum albumin (AFB1 −BSA) conjugate. Upon addition of target AFB1, the analyte initially competes with AFB1−BSA-GOx for the labeled anti-AFB1 antibody on magnetic bead, and then the carried GOx oxidizes the glucose to generate H2O2. In this case, the as-produced H 2 O 2 can be used as an efficient scavenger for the decomposition of MnO2 nanosheets (MnO2 + H2O2 + 2H+ → Mn2+ + 2H2O + O2). Because of the decline of MnO2CQDs on the electrode, the photocurrent of MnO2-CQDsmodified electrode decreases with the increasing H2 O 2 concentration. By monitoring the change in photocurrent, we can quantitatively determine AFB1 in the sample. Because MnO2-CQDs were eliminated, favorably, a complementary

detection method. Therefore, further research is necessary to improve the efficiency and practicality of the PEC system, particularly the development of high-efficiency and costeffective materials. Manganese oxide (MnO2, a compound of low-cost and high activity in neutral or alkaline media) has been used widely in the electrochemical sensing system due to its high specific surface area, environmental friendliness, abundant availability, strong adsorption ability, and easy preparation.16,17 MnO2 nanosheets with a band gap of about 2.1 eV determined from UV−vis spectropy18 have an absorption peak centered at around 380 nm, which is tailed to the visible regime.19,20 Sakai et al. reported the observation of photocurrent generation via self-assembled MnO2 nanosheets on indium tin oxide (ITO) under visible light irradiation in nonaqueous electrolyte.21 It is unfortunate to observe the poor photocurrent of MnO2 nanosheets, perhaps due to the following two reasons: a strongly localized d−d transition and a high contact resistance. The layered number of MnO2 is on account of the recombination loss in exciting electron−hole pairs, while the latter one decreases carriers during transport.18−20 To overcome the shortcoming, the emergence of hybrid nanostructures opens a new horizon for improvement of photoelectrochemical properties.22−25 Carbon quantum dots (CQDs) are considered to be organic quantum dots (QDs) synthesized with a nontoxic, simple, and inexpensive route, and have demonstrated high and stable photoluminescence, excellent water solubility, high resistance to photobleaching, good biocompatibility, and low toxicity.26,27 Especially, their small particle size, rapid electron transfer, and excellent conductivity make it easier to form composites and thus further strengthen the PEC properties of the original materials.26,28 MnO2 coupling with CQDs conducts a suitable composite material to improve the PEC activity of MnO2 because the favorable conductivity of CQDs can efficiently improve the charge-separation rate of MnO2. Simultaneously, the small-size CQDs and highly transparent morphology of MnO2 nanosheets would increase the light 5638

DOI: 10.1021/acs.analchem.7b00942 Anal. Chem. 2017, 89, 5637−5645

Article

Analytical Chemistry

ultrapure water three times; and (iii) the as-prepared precipitate was redispersed into 1.0 mL of ultrapure water for the following usage. Conjugation of Magnetic Beads with Monoclonal Anti-AFB1 Antibody (MB-mAb). In this work, modification of the carboxylated magnetic beads with monoclonal anti-AFB1 antibody was referred to in our previous reports by a typical carbodiimide coupling.35,36 Briefly, the carboxylated magnetic beads (1.0 mL, 5.0 mg mL−1) were initially blended with MES solution (0.1 M, 1.0 mL, pH 5.2) containing 400 mM EDC and 100 mM NHS, and the resulting suspension was gently shaken on a shaker for 30 min at rt (note: the aim of this step is to activate the −COOH group on magnetic beads). The functional magnetic beads were magnetically separated and washed with ultrapure water three times to remove the excess EDC and NHS. Following that, the collected magnetic beads were dispersed into monoclonal anti-AFB1 antibody solution (1.0 mL, 500 μg mL−1), and reacted for 12 h at 4 °C with slight shaking. After completion of the incubation, the conjugate (i.e., MB-mAb) was collected by using an external magnet and washed with PBS (0.01 M, pH 7.4). Finally, the MB-mAb conjugate was dispersed in 1.0 mL of PBS (0.01 M, pH 7.4) containing 1.0 wt % BSA, and stored at 4 °C when not in use. Conjugate of AFB1−BSA with Glucose Oxidase (AFB1− BSA-GOx). AFB1−BSA conjugates were functionalized with glucose oxidase similar to our previous report.37 GOx (200 μL, 500 μg mL−1) and AFB1−BSA (50 μL, 500 μg mL−1) were initially dispersed into 1.0 mL of PBS (10 mM, pH 7.4), and the pH was adjusted to 9.5 by 1.0 M K2CO3. Following that, glutaraldehyde (0.1 mL, original concentration) was injected to the mixture, and incubated for 12 h at 4 °C with gentle shaking. After that, the pH was adjusted again to 7.4 by using NaH2PO4, and the obtained mixture was purified by ultrafiltration 12−15 times to remove the unbound GOx and AFB1−BSA. Finally, the obtained AFB1−BSA-GOx conjugates were dispersed into 1.0 mL of PBS (0.1 M, pH 7.4). Magneto-Controlled Immunoreaction Protocol and Photoelectrochemical Measurement. Magneto-controlled microfluidic device was designed and schematically illustrated in our recent report.38 The device mainly consists of a semiautomatic flow-through system and a PEC detection cell [note: the cell included a movable fluorine-doped tin oxide (FTO) electrode at the bottom, an external permanent BaFe12O19 magnet under the FTO electrode, and a 500 W Xe lamp with a 420 nm cutoff filter above]. A six-port rotary valve installed with a 1.0 mL syringe pump was connected to the detection cell by a Teflon tube. Prior to the experiment, FTO electrode was washed by sonication in acetone and distilled water in sequence, and then blow-dried with nitrogen. Following that, the waterproof transparent tape punched with a hole (r = 2.5 mm) was stuck on the surface of FTO electrode, and the active area was ∼19.625 mm2. Afterward, 30 μL of the above-prepared MnO2-CQDs was dropped on the surface of FTO electrode, and dried at rt. Subsequently, the modified FTO electrode was installed in the detection cell for the detection of target AFB1 as follows: (i) the as-prepared MBmAb (100 μL) was flowed through the detection cell and collected on the electrode with the aid of an external magnet; (ii) the different-concentration AFB1 standard/sample (50 μL) (note: these standards or samples were diluted with 0.1 M PBS, pH 7.4) and the as-prepared AFB1−BSA-GOx (50 μL) were simultaneously injected in the detection cell (detaching the magnet), and reacted for 60 min at rt to execute the

visual detection can be realized by judging the color intensity change of the electrode. The aim of this work is to successfully design a new semiautomatic PEC detection platform for quantitative detection of mycotoxins.

■

EXPERIMENTAL SECTION Materials and Chemicals. Monoclonal anti-AFB1 antibody (from School of Food Science and Technology, Jiangnan University, Wuxi, China), AFB1−BSA conjugate (from Oil Crops Research Institute of the Chinese Academy of Agricultural Sciences, Wuhan, China), and the standards of aflatoxins including AFB1 , AFB2, AFG1, and AFG2 in acetonitrile (from Express Technology Co. Ltd., Beijing, China) were used in this work. 1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC), N-hydroxysuccinimide (NHS), glucose oxidase (GOx) from Aspergillus niger, tetramethylammonium hydroxide pentahydrate (TMA·OH), and glucose were purchased from Sigma-Aldrich (St. Louis, MO). Manganese chloride tetrahydrate (MnCl2·4H2O), hydrogen peroxide (H2O 2, concn = 30 wt %), citric acid monohydrate (C6H8O7·4H2O), and urea [CO(NH2)2] were achieved from Sinopharm Chem. Re. Co. Ltd. (Shanghai, China). Carboxylated magnetic beads (MBs, particle size: ∼100 nm) in an aqueous suspension with a concentration of 25 mg mL −1 were obtained from Chemicell GmbH (Berlin, Germany). All other chemicals were of analytical grade without further purification. Ultrapure water (18.2 MΩ cm−1) was prepared via a Millipore water purification system (Milli-Q, Merck KGaA, Germany). All buffers including the phosphatebuffered saline (PBS) solution were the products of SigmaAldrich. Synthesis of Carbon Quantum Dots-Functionalized MnO2 Nanosheets. MnO2 nanosheets coated with carbon quantum dots (MnO2-CQDs) were synthesized by the wetchemistry method. The first stage is to prepare MnO 2 nanosheets according to the literature.33 Briefly, the solution (20 mL) containing 0.6 M TMA·OH and 3.0 wt % H2O2 was initially added into MnCl2 aqueous solution (10 mL, 0.3 M), and the resulting mixture was then stirred vigorously for 12 h at room temperature (rt) in the open air to obtain a dark brown suspension. Afterward, the suspension was centrifuged (12 000g, 10 min), and the obtained crude product was washed with ethanol and water three times. To remove the residual solvent, the precipitate was dried under high vacuum at 60 °C for 12 h for further use. Next, nitrogen-doped carbon quantum dots (CQDs) were synthesized via a typical microwave synthesis method.34 Initially, citric acid monohydrate (6.0 g) and urea (6.0 g) were dispersed into 20 mL of ultrapure water, and heated in a domestic 800 W microwave oven for ∼4 min. During this process, the solution changed from a colorless liquid to darkbrown clustered solid. Following that, the remaining small molecules in the product and the agglomerated/large particles were removed by drying in a vacuum oven (60 °C, 60 min) and centrifugation (3 000g, 15 min) , respectively. Finally, the obtained CQDs were utilized for the preparation of MnO2CQDs as follows: (i) MnO2 nanosheets with different masses (from 0 to 100 μg) were dispensed into CQDs aqueous solution (1.0 mL, 2.4 μg mL−1), respectively, and the mixture was adequately shaken on a shaker overnight (note: CQDs were attached onto the surface of MnO2 nanosheets via physical adsorption during this process); (ii) the resultant suspension was centrifuged (12 000g, 10 min) and washed with 5639

DOI: 10.1021/acs.analchem.7b00942 Anal. Chem. 2017, 89, 5637−5645

Article

Analytical Chemistry

Figure 1. (A,B) HRTEM images of (A) MnO2 nanosheets and (B) MnO2-CQDs (insets: the corresponding magnification images); (C) Raman patterns of (a) MnO2 nanosheets and (b) MnO2-CQDs; and (D) FTIR spectra, (E) UV−vis absorption spectra, and (F) fluorescence spectra of (a) MnO2 nanosheets, (b) CQDs, and (c) MnO2-CQDs (insets: the corresponding photographs).

the symmetric stretching vibration, respectively.39 Unfavorably, we could not clearly observe Raman characteristic peaks for CQDs (curve b). The reason might be the fact that the peaks of MnO2 nanosheets interfered with those of CQDs. To further demonstrate the existence of CQDs on the nanosheets, Fourier transform infrared (FTIR) spectroscopy (PerkinElmer, U.S.) was utilized because the nanohybrids were constructed through the electrostatic interaction of negatively charged MnO2 nanosheets and positively charged CQDs (Figure 1D). Pure CQDs exhibited three characteristic peaks at 3400, 1750, and 1607 cm−1 corresponding to the O−H, CO, and N−H, respectively (curve b).40,41 In contrast, the FTIR spectrum of MnO2 nanosheets had two peaks at 520 and 710 cm−1 assigned to the stretching vibrations of Mn−O and Mn−O−Mn bonds, respectively, while another peak at 3400 cm−1 was ascribed to the vibration of O−H in the nanosheets (curve a).42 Significantly, these characteristic peaks for CQD and MnO2 simultaneously appeared after interaction of MnO2 nanosheets with CQDs (curve c). This phenomenon could be also verified by UV−vis absorption spectroscopy (Tecan Infinite 200 Pro, TECAN, Switzerland). As shown in Figure 1E, MnO 2 nanosheets exhibited a wide band in the range of 300−600 nm and a peak located at 370 nm (curve a), while two absorbance bands for CQDs were located at around 330 and 420 nm (curve b). These two characteristic absorbance bands for CQDs mainly derived from the trapping of excited-state energy of the surface states, which was conducive for strong fluorescence intensity (Figure 1F, curve b, photogram b).43 Moreover, the characteristic peak at 330 nm could be also found at MnO2-CQDs (curve c). In addition, we also observed that the fluorescence of the CQDs was quenched after the formation of MnO2-CQDs because of the resonance energy transfer between two nanostructures (Figure 1F). The results further revealed that MnO2-CQDs could be successfully synthesized by using our designed route. Characteristics of Photoelectrochemical Immunosensing Platform. For the development of the MnO2-

competitive-type immunoreaction; (iii) after washing with pH 7.4 PBS (attaching the magnet), glucose substrate (100 μL, 20 mM) in PBS (0.1 M, pH 5.5) was flowed into the cell and reacted for 15 min; and (iv) the photoelectrochemical measurement was registered and recorded at the applied potential of 500 mV on an AutoLab electrochemical workstation (μAutIII.Fra2.v, Eco Chemie, Netherlands) with a classical three-electrode system including a MnO2-CQDsmodified FTO working electrode, a Pt-wire counter electrode, and an Ag/AgCl reference electrode. All determinations were done at least in duplicate.

■

RESULTS AND DISCUSSION Characterization of MnO2-CQDs Photoactive Nanomaterials. To develop a highly efficient photoelectrochemical sensing platform, the successful preparation of photoactive materials was very crucial. First, we used a high-solution transmission electron microscope (HRTEM, H-7650, Hitachi, Japan) to characterize the morphology and surface structure of MnO2-CQDs. As shown in Figure 1A, the as-prepared samples displayed a typical planar sheet-like morphology with an irregular shape, indicating that nanostructures were neither MnO2 nanoparticles nor the powder. Moreover, we could also observe that MnO2 nanosheets had a high-degree of transparency with an ultrathin nature (Figure 1A, inset). The dark strips corresponded to the folded edges or wrinkles of MnO2 nanosheets. Such a sheet-like structure could provide a large surface coverage for the immobilization of carbon quantum dots. When MnO2 nanosheets interacted with CQDs, we obviously found that numerous nanoparticles were coated on the surface of nanosheets (Figure 1B). The lattice and shape of CQDs appeared, and the average size was 3−5 nm (Figure 1B, inset). Furthermore, the as-synthesized MnO2-CQDs were monitored by using Raman spectroscopy (RFS, Bruker, Germany) (Figure 1C). Two sharp characteristic peaks for MnO2 nanosheets at 572 and 645 cm−1 (curve a) were acquired, which was ascribed to Mn−O stretching vibration and 5640

DOI: 10.1021/acs.analchem.7b00942 Anal. Chem. 2017, 89, 5637−5645

Article

Analytical Chemistry

Figure 2. (A) Fluorescent spectra of (a) MnO2-CQDs, (b) MnO2-CQDs + H2O2, (c) MnO2-CQDs + GOx, (d) MnO2-CQDs + glucose, and (e) MnO2-CQDs + GOx + glucose; (B) electrochemical impedance spectroscopy (inset: equivalent circuit) and (C) photoelectrochemical responses for (a) FTO, (b) MnO2/FTO, (c) CQDs/FTO, and (d) MnO2-CQDs/FTO; and (D) photocurrent responses of MnO2-CQDs/FTO in the (a) absence and (b) presence of H2O2 (note: Nyqusit diagrams were obtained in 5 mM Fe(CN)64−/3− + 0.1 M KCl with the range from 10−2 to 105 Hz at an alternate voltage of 5 mV, while the photocurrents were obtained in 0.1 M PBS of pH 5.5).

Logically, a puzzling question arises as to whether CQDs could enhance the photocurrent of MnO2 nanosheets after formation of MnO2-CQDs. Initially, we used electrochemical impedance spectroscopy (EIS) to investigate the conductivity of MnO2 nanosheets before and after coating with CQDs because it is usually utilized for understanding chemical formations/processes associated with the conductive supports (Figure 2B). These EIS data were fitted to a Randles equivalent circuit (Figure 2B, inset), which contains electrolyte resistance (Rs), the lipid bilayer capacitance (Cdl), charge transfer resistance (Ret), and the Warburg element (Zw). Often, the semicircular diameter in the Nyquist plots of EIS is equal to the value of electron transfer resistance (Ret). Relative to bare FTO electrode (curve a), modification of MnO2 nanosheets or CQDs caused the increasing resistance of FTO electrode (curves b,c), suggesting that introduction of nanomaterials hindered the electron transfer. Significantly, the resistance of MnO2-CQDs/FTO was obviously lower than that of MnO2/ FTO (curve d vs curve b). Thus, CQDs could improve the conductivity of MnO2 nanosheets. Undoubtedly, another concern on the nanohybrids is whether the photocurrent of MnO2-CQDs could be improved under visible light illumination in comparison with MnO2 or CQDs alone. As shown in Figure 2C, almost no photocurrents were achieved at bare FTO (curve a) and CQDs/FTO (curve c), and a relatively weak photocurrent was observed at MnO2/FTO (curve b). When MnO2-CQDs were modified on the FTO electrode, however, a strong photocurrent appeared (curve d), indicating that MnO2 nanosheets coated with CQDs could accelerate the electron transfer rate, and decrease the recombination efficiency of the photogenerated electron and hole of MnO2 nanosheets. Furthermore, the photocurrent intensity of MnO2-CQDs/

CQDs-based PEC immunosensing platform, a prerequisite lies in the fact that the catalytic product (H2O2) of GOx toward glucose could readily reduce MnO2 nanosheets into Mn2+. To clarify this issue, the fluorescence intensities of MnO2-CQDs were investigated under different conditions (Figure 2A). Similar to the above-mentioned result (Figure 1F, curve c), the fluorescence signal was almost observed at MnO2-CQDs (Figure 2A, curve a). Upon addition of excess H2O2 into the MnO2-CQDs, a strong fluorescence signal appeared (curve b), indicating that CQDs were dissociated from the nanosheets. The reason should be ascribed to the dissolution of MnO2 nanosheets in the presence of H2O2 via the redox reaction. Typically, H2O2 can be acquired by GOx toward the catalytic oxidization of glucose. As control tests, we first monitored whether GOx and glucose alone could cause the dissolution of CQDs from the nanosheets. As seen from curves c and d, the fluorescence signals could be almost the same as that of pure MnO2-CQDs (curve a). Favorably, the fluorescence signal largely increased in the simultaneous presence of GOx and glucose due to H2O2 formation (curve e). The formed H2O2 reduced MnO2 into Mn2+, thus resulting in the decomposition of MnO2 nanosheets and the releasing of CQDs. Relative to curve b by using excess H2O2, the GOx−glucose system caused stronger fluorescence intensity (curve e). The reason might be most likely as a consequence of the fact that excess H2O2 could quench the fluorescence of CQDs.44,45 These results are in accordance with those of the corresponding photographs (Figure 2A, insets), further indicating that the GOx−glucose system could be used for the dissociation of CQDs from the MnO2-CQDs by the as-produced H2O2. In this case, GOx can be employed as the label for the development of enzyme immunoassays. 5641

DOI: 10.1021/acs.analchem.7b00942 Anal. Chem. 2017, 89, 5637−5645

Article

Analytical Chemistry

Figure 3. Effects of (A) MnO2 concentration for preparation of MnO2-CQDs, (B) the applied potential for PEC measurement, and (C) the reaction time between MnO2-CQDs and the as-produced H2O2 on the photocurrent of the modified electrode in 0.1 M PBS (pH 5.5) (note: 5.0 ng mL−1 AFB1 used in the A and C cases).

Figure 4. (A) Photocurrent responses of MnO2-CQDs/FTO in 0.1 M PBS (pH 5.5) containing 20 mM glucose after reaction of magnetic immunoassay system with different-concentration target AFB1 standards (inset: the corresponding photograph of MnO2-CQDs/FTO after measurement), (B) calibration curve, (C) the stability of the MnO2-CQDs/FTO, and (D) the specificity of this immunosensing platform against AFB1 (0.1 ng mL−1), AFB2 (100 ng mL−1), AFG1 (100 ng mL−1), AFG2 (100 ng mL−1), BTB (100 ng mL−1), OTA (100 ng mL−1), and OA (100 ng mL−1) in PBS (0.1 M, pH 7.4).

before and after reaction with 5.0 ng mL−1 AFB1 (used as an example), coupling with the magneto-controlled microfluidic device. As seen from Figure 3A, the change in the photocurrent initially increased with the increasing MnO2 concentration, and then decreased. The maximum shift in the photocurrent was achieved at 60 μg mL−1 MnO2 nanosheets. Hence, 60 μg mL−1 MnO2 nanosheets was utilized for the preparation of MnO2CQDs in 1.0 mL of CQDs (2.4 μg mL−1). At this condition, we also investigated the effect of the applied potential on the photocurrent of the newly prepared MnO2-CQDs/FTO alone (i.e., in the absence of magnetic immunoassay). The optimal photocurrent was acquired at 0.5 V (Figure 3B). A higher or lower potential would decrease the photocurrent of MnO2CQDs/FTO. So, 0.5 V was chosen for PEC measurement in this study. Certainly, the reaction time between MnO2-CQDs and the immunoreaction product (H2O2) also greatly affects the

FTO largely decreased upon H2O2 introduction in this system (Figure 2D, curve b vs curve a). On the basis of the abovementioned results, we might conclude that MnO2-CQDs could be used for the development of enzyme immunoassay by using the GOx-labeled strategy. Optimization of Experimental Conditions. As described above, the detectable signal was derived from the produced photocurrent of MnO2-CQDs. Generally speaking, the doping ratio between MnO2 nanosheets and CQDs during this preparation would directly affect the photocurrent signal. Although a high-concentration CQD in the nanohybrids could improve the conductivity of the MnO2-CQDs/FTO, it might decrease the photocurrent response. The comparative study was carried out by synthesizing a series of MnO2-CQDs with different MnO2 concentrations from 0 to 100 μg mL−1 in 1.0 mL of CQDs (2.4 μg mL−1). The evaluation was based on the change (ΔI) in the photocurrent of MnO2-CQDs/FTO 5642

DOI: 10.1021/acs.analchem.7b00942 Anal. Chem. 2017, 89, 5637−5645

Article

Analytical Chemistry

Table 1. Comparison of the Results from MnO2-CQDs-Based PEC Immunoassay and Commercial AFB1 ELISA Kit (Diagnostic Automation Inc., LOD: 5.0 pg mL−1) for Naturally Contaminated and Spiked Peanut Samples for Evaluation of Method Accuracy method; concn [mean ± SD (RSD), ng mL−1, n = 3]a typeb spiked peanut

naturally contaminated peanut

no. 1 2 3 4 5 6 7 8 9 10 11 12

PEC immunoassay 15.22 1.28 7.32 23.22 45.53 0.27 0.23 0.89 15.36 9.53 32.89 5.56

± ± ± ± ± ± ± ± ± ± ± ±

1.23 0.05 0.53 1.98 2.05 0.02 0.01 0.12 1.25 0.86 2.42 0.31

(8.08%) (3.91%) (7.24%) (8.53%) (4.50%) (7.41%) (4.35%) (13.48%) (8.14%) (9.02%) (7.36%) (5.58%)

texptl

AFB1 ELISA kit 15.80 1.35 7.08 21.39 43.05 0.26 0.25 0.95 14.82 9.78 34.91 5.88

± ± ± ± ± ± ± ± ± ± ± ±

1.12 0.08 0.52 1.83 2.23 0.02 0.02 0.11 1.32 0.92 2.53 0.35

(7.09%) (5.93%) (7.34%) (8.56%) (5.18%) (7.69%) (8.70%) (11.57%) (8.91%) (9.41%) (7.25%) (5.95%)

0.60 1.29 0.56 1.18 1.42 0.61 1.55 0.64 1.47 0.34 1.00 1.19

The regression equation (linear) for these data from the average values (mean) between two methods was as follows: y = (0.97 ± 0.02)x + (0.21 ± 0.44) (r = 0.997, n = 12; x axis, by MnO2-CQDs-based PEC immunoassay; y axis, by AFB1 ELISA kit). All data as mean ± SD were obtained on the basis of three measurements. The high-concentration AFB1 samples (nos. 4, 5, and 11) were calculated on account of the dilution ratio during the measurement. bSamples 1−6 were spiked with AFB1 standards in blank peanut, while samples 6−12 were naturally contaminated. a

(e.g., LOD, 100 ppt for Quicking Biotech Inc.; LOD, 5.0 pg mL−1 for Diagnostic Automation Inc.; 50 pg mL−1 for MaxSignal Inc.; 250 pg mL−1 for MyBioSource Inc.). Such a low LOD could be suitable for the detection of lowconcentration AFB1. The reproducibility and stability of MnO2-CQDs-based PEC immunoassay were important for the development of a new analytical method. Initially, we investigated the reproducibility of our strategy by assaying three AFB1 levels (low-middle-high) in buffer samples containing 0.05, 0.5, and 10 ng mL−1. The results were evaluated by calculating the intra- and interbatch relative standard deviation (RSD). Experimental data indicated that the RSDs were 6.7%, 4.3%, and 7.4% for intra-assay, and 10.7%, 9.5%, and 8.9% for inter-assay toward the abovementioned concentrations, respectively (n = 3). Meanwhile, we monitored the stability of the MnO2-CQDs- modified FTO electrode under the on−off light irradiation during a period of 400 s (Figure 4C). All photocurrents during this process were almost reproducible (RSD = 11.6%, n = 18). So, the MnO2CQDs-based PEC immunosensing system could be used repeatedly, and further verified the possibility of batch preparation. Next, we investigated the specificity of MnO2-CQDs-based PEC immunoassay by challenging this system against other mycotoxins or marine toxins, for example, AFB2, AFG1, AFG2, ochratoxin A (OTA), brevetoxin B (BTB), and okadaic acid (OA). As shown in Figure 4D, nontarget analytes including AFG1, AFG2, OTA, BTB, and OA only caused a similar photocurrent with the blank sample. Unfavorably, AFB2 exhibited a relatively high cross-reactivity with this system, which was ascribed to their similar molecular structures. A mixture of target AFB1 with nontargets did not cause the significant change in the photocurrent relative to AFB1 alone, except for AFB2. Thus, the developed PEC immunoassay could be identified by different-structure small molecules. Monitoring of Spiked or Naturally Contaminated Peanut Samples. To fulfill the future application of the newly developed method, one precondition should meet the needs of real samples in the complex system. To this end, naturally contaminated and spiked peanut samples were

photocurrent of this sensing platform. A short reaction time is unfavorable for the generation of H2O2 by GOx-catalyzed glucose and dissolution/etching of MnO2 nanosheets. As shown in Figure 3C, the photocurrent of the MnO2-CQDs/ FTO decreased with the increasing reaction time, and tended to level off after 15 min. A longer reaction time did not significantly improve the photocurrent signal. To save the assay time, 15 min was selected for the enzymatic reaction and used as the etching time for MnO2-CQDs by the produced H2O2. Analytical Performance of MnO2-CQDs-Based Immunosensing Platform toward AFB1. By coupling with photoelectrochemical properties of MnO2-CQDs and the H2O2−MnO2 etching system, a novel signal-on competitivetype immunoassay was designed for the detection of smallmolecular AFB1 on anti-AFB1 antibody-conjugated magnetic beads, using GOx-labeled BSA-AFB1 conjugate as the tag. Accompanying the formation of immunocomplexes on magnetic beads, the carried GOx oxidized glucose to gluconic acid and H2O2. The generated H2O2 reacted with MnO2 to dissolve or etch the nanosheets, thus resulting in the dissociation of CQDs from the electrode and decreasing the photocurrent of MnO2-CQDs/FTO. Meanwhile, the decomposition of MnO2 nanosheets bleached the color of the modified FTO electrode out. As shown in Figure 4A, the photocurrent of the MnO2-CQDs-based competitive-type immunosensing platform increased with increasing target AFB1 concentration. A low-concentration AFB1 caused a very weak photocurrent and a relatively light color on the electrode (Figure 4A, inset). With the increasing AFB1 level, the color gradually became thick, which could be distinguished with the naked eye. By the change in the color, we could qualitatively evaluate the concentration of target AFB1 in the sample. In addition, a good linear relationship was achieved between the photocurrent and AFB1 concentration in the dynamic range of 0.01−20 ng mL−1 with a limit of detection of 2.1 pg mL−1 (ppt) at a signal-to-noise ratio of 3 (Figure 4B). The linear regression equation could be fitted to y (μA) = 0.259 × C[AFB1] + 0.568 (ng mL−1, R2 = 0.998, n = 8), which could quantitatively calculate the exact level of AFB1. Significantly, the LOD of our system was lower than those of currently used AFB1 ELISA kits 5643

DOI: 10.1021/acs.analchem.7b00942 Anal. Chem. 2017, 89, 5637−5645

Article

Analytical Chemistry prepared on the basis of our previous reports.31,32 Thereafter, these as-prepared samples were determined by using a MnO2CQDs-based PEC immunoassay and AFB1 ELISA Kit purchased from Diagnostic Automation Inc., respectively. Comparison of the results obtained from these two methods was performed by using a t-test and a least-squares regression method. As indicated from Table 1, the regression equation for these data from the average values between two methods was fitted to y = (0.97 ± 0.02)x + (0.21 ± 0.44) (r = 0.997, n = 12; x axis, MnO2-CQDs-based PEC immunoassay; y axis, AFB1 ELISA kit). The standard deviations of the slope and intercept46 were given in the regression equation. The correlation between two methods was evaluated via use of ttests for comparing the experimental values of the intercept and slope with the ideal situation of zero intercept and slope of 1. The statistics t for the slope and intercept were calculated according to the following equation: t = (a − 0)/sa and t = (b − 1)/sb, where a and b stand for the intercept and slope, respectively, and sa and sb for standard deviation of the intercept and slope, respectively. No significance differences at the 0.05 significance level were encountered between the optimum values of intercept and slope and experimental data. Importantly, all texptl values in these samples were lower than tcrit (tcrit[0.05,4] = 2.77), indicating a good agreement between two methods. Hence, the accuracy of this method was acceptable, and could be preliminarily applied for the detection of target AFB1 in the complex samples.

Dietmar Knopp: 0000-0003-4566-9798 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was financially supported by the National Natural Science Foundation of China (21475025 and 21675029), the National Science Foundation of Fujian Province (2014J07001), the Program for Changjiang Scholars and Innovative Research Team in University (IRT15R11), and the Alexander von Humboldt Foundation of Germany.

■

■

CONCLUSIONS In summary, this work demonstrated a novel magnetocontrolled immunosensing system for the simultaneous PEC and visual detection of small-molecular mycotoxins (AFB1 used in this case) by coupling a high-throughput microfluidic device with nanoparticles-based immunoassay format. Introduction of high-efficient photosensitive materials, MnO2-CQDs, enhanced the sensitivity of the PEC immunoassay. As a typical catalytic product of GOx, H2O2-repsonsive dissolution/etching of MnO2 nanosheets innovated the conventional competitive-type immunoassay from the signal-off format to the signal-on mode, and thereby was suitable for the detection of small molecules. Because the as-synthesized MnO2 nanosheets could quench the fluorescence signal of CQDs, this system can be applied for fluorescence sensing of target analytes. Further, CQDs could be readily dissociated from the nanosheets by the product H2O2. By changing carbon dots with other materials, for example, CdS quantum dots or methylene blue, the detection scheme can be extended using other devices. In addition, MnO2-CQDs could be realized by using the naked eye, accompanying qualitative evaluation with the change in the color and the quantitative determination by the linear regression equation. Although the present system focused on the detection of target AFB1, it easily extends to determine other mycotoxins by controlling the corresponding antibody.

■

REFERENCES

(1) Nascimento, C.; Santos, P.; Pereira-Filho, E.; Rocha, F. Food Chem. 2017, 221, 1232−1244. (2) Gil-Santos, E.; Baker, C.; Lemaitre, A.; Gomez, C.; Leo, G.; Favero, I. Nat. Commun. 2017, 8, 14267. (3) Hou, Y.; Qiu, M.; Zhang, T.; Ma, J.; Liu, S.; Zhuang, X.; Yuan, C.; Feng, X. Adv. Mater. 2017, 29, 1604480. (4) Li, C.; Li, A.; Luo, Z.; Zhang, J.; Chang, X.; Huang, Z.; Wang, T.; Gong, J. Angew. Chem., Int. Ed. 2017, 56, 4150. (5) Pardo-Yissar, V.; Katz, E.; Wasserman, J.; Willner, I. J. Am. Chem. Soc. 2003, 125, 622−623. (6) Tu, W.; Cao, H.; Zhang, L.; Bao, J.; Liu, X.; Dai, Z. Anal. Chem. 2016, 88, 10459−10465. (7) Li, L.; Zhang, Y.; Zhang, L.; Ge, S.; Liu, H.; Ren, N.; Yan, M.; Yu, J. Anal. Chem. 2016, 88, 5369−5377. (8) Ge, L.; Wang, P.; Ge, S.; Li, N.; Yu, J.; Yan, M.; Huang, J. Anal. Chem. 2013, 85, 3961−3970. (9) Fan, G.; Shi, X.; Zhang, J.; Zhu, J. Anal. Chem. 2016, 88, 10352− 10356. (10) Wang, Y.; Zhao, X.; Tian, Y.; Wang, Y.; Jan, A.; Chen, Y. Chem. Eur. J. 2017, 23, 419−426. (11) Seabold, J.; Shankar, K.; Wike, R.; Paulose, M.; Varghese, O.; Grimes, G.; Choi, K. Chem. Mater. 2008, 20, 5266−5273. (12) Zhuang, J.; Lai, W.; Xu, M.; Zhou, Q.; Tang, D. ACS Appl. Mater. Interfaces 2015, 7, 8330−8338. (13) Lin, Y.; Zhou, Q.; Tang, D.; Niessner, R.; Yang, H.; Knopp, D. Anal. Chem. 2016, 88, 7858−7866. (14) Zhuang, J.; Tang, D.; Lai, W.; Xu, M.; Tang, D. Anal. Chem. 2015, 87, 9473−9480. (15) Fan, C.; Han, L.; Zhu, H.; Zhang, J.; Zhu, J. Anal. Chem. 2014, 86, 12398−12405. (16) Najafpour, M.; Renger, G.; Holynska, M.; Moghaddam, A.; Aro, E.; Carpentier, R.; Nisshihara, H.; Eaton-Rye, J.; Shen, J.; Allakhverdiev, S. Chem. Rev. 2016, 116, 2886−2936. (17) Chen, Z.; Jiao, Z.; Pan, D.; Li, Z.; Wu, M.; Shek, C.; Wu, C.; Lai, J. Chem. Rev. 2012, 112, 3833−3855. (18) Fan, H.; Yan, G.; Zhao, Z.; Hu, X.; Zhang, W.; Liu, H.; Fu, X.; Fu, T.; Zhang, X.; Tan, W. Angew. Chem., Int. Ed. 2016, 55, 5477− 5482. (19) Kwon, K.; Refson, K.; Sposito, G. Phys. Rev. Lett. 2008, 100, 146601. (20) Marafatto, F.; Strader, M.; Gonzalez-Holgueral, J.; Schwartzberg, A.; Gilbert, B.; Pena, J. Proc. Natl. Acad. Sci. U. S. A. 2015, 112, 4600− 4605. (21) Sasai, K.; Ebina, Y.; Takada, K.; Sasaki, T. J. Phys. Chem. B 2005, 109, 9651−9655. (22) Dai, H.; Zhang, S.; Hong, Z.; Lin, Y. Anal. Chem. 2016, 88, 9532−9538. (23) Zhang, Y.; Ge, L.; Li, M.; Ge, S.; Yu, J.; Song, X.; Cao, B. Chem. Commun. 2014, 50, 1417−1419. (24) Fan, L.; Zhao, G.; Shi, H.; Liu, M.; Wang, Y.; Ke, H. Environ. Sci. Technol. 2014, 48, 5754−5761. (25) Zhang, Y.; Yan, M.; Ge, S.; Ma, C.; Yu, J.; Song, X. J. Mater. Chem. B 2016, 4, 4980−4987.

AUTHOR INFORMATION

Corresponding Authors

*Phone: +86-591-2286 6125. Fax: +86-591-2286 6135. E-mail: [email protected]. *Phone: +49-89-2180 78252. Fax: +49-89-2180 78255. E-mail: [email protected]. ORCID

Dianping Tang: 0000-0002-0134-3983 5644

DOI: 10.1021/acs.analchem.7b00942 Anal. Chem. 2017, 89, 5637−5645

Article

Analytical Chemistry (26) Yu, H.; Shi, R.; Zhao, Y.; Waterhouse, G.; Wu, L.; Tung, C.; Zhang, T. Adv. Mater. 2016, 28, 9454−9477. (27) Lim, S.; Shen, W.; Gao, Z. Chem. Soc. Rev. 2015, 44, 362−381. (28) Shi, W.; Zhang, X.; Brillet, J.; Huang, D.; Li, M.; Wang, M.; Shen, Y. Carbon 2016, 105, 387−393. (29) Fan, W.; Bu, W.; Shen, B.; He, Q.; Cui, Z.; Liu, Y.; Zheng, X.; Zhao, K.; Shi, J. Adv. Mater. 2015, 27, 4155−4161. (30) Yuan, J.; Cen, Y.; Kong, X.; Wu, S.; Liu, C.; Yu, R.; Chu, X. ACS Appl. Mater. Interfaces 2015, 7, 10548−10555. (31) Lin, Y.; Zhou, Q.; Lin, Y.; Tang, D.; Niessner, R.; Knopp, D. Anal. Chem. 2015, 87, 8531−8540. (32) Tang, D.; Lin, Y.; Zhou, Q.; Lin, Y.; Li, P.; Niessner, R.; Knopp, D. Anal. Chem. 2014, 86, 11451−11458. (33) Kai, K.; Yoshida, Y.; Kageyama, H.; Saito, G.; Ishigaki, T.; Furukawa, Y.; Kawamata, J. J. Am. Chem. Soc. 2008, 130, 15938− 15943. (34) Qu, S.; Wang, X.; Lu, Q.; Liu, X.; Wang, L. Angew. Chem., Int. Ed. 2012, 51, 12215−12218. (35) Zhang, B.; Liu, B.; Tang, D.; Niessner, R.; Chen, G.; Knopp, D. Anal. Chem. 2012, 84, 5392−5399. (36) Zhang, B.; Tang, D.; Goryacheva, I.; Niessner, R.; Knopp, D. Chem. - Eur. J. 2013, 19, 2496−2503. (37) Tang, D.; Su, B.; Tang, J.; Ren, J.; Chen, G. Anal. Chem. 2010, 82, 1572−1534. (38) Shu, J.; Qiu, Z.; Lin, Z.; Cai, G.; Yang, H.; Tang, D. Anal. Chem. 2016, 88, 12539−12546. (39) Julien, C.; Massot, M.; Rangan, S.; Lemal, M.; Guyomard, D. J. Raman Spectrosc. 2002, 33, 223−228. (40) Arul, V.; Edison, T.; Lee, Y.; Sethuraman, M. J. Photochem. Photobiol., B 2017, 168, 142−148. (41) Zhang, J.; Chen, Y.; Tan, J.; Sang, H.; Zhang, L.; Yue, D. Appl. Surf. Sci. 2017, 396, 1138−1145. (42) Mallakpour, S.; Motirasoul, F. Mater. Chem. Phys. 2017, 191, 188−196. (43) Wang, Y.; Jiang, X. Sci. China: Chem. 2016, 59, 836−842. (44) Lan, M.; Di, Y.; Zhu, X.; Ng, T.; Xia, J.; Liu, W.; Meng, X.; Wang, P.; Lee, C.; Zhang, W. Chem. Commun. 2015, 51, 15574− 15577. (45) Shan, X.; Chai, L.; Ma, J.; Qian, Z.; Chen, J.; Feng, H. Analyst 2014, 139, 2322−2325. (46) Miller, J. N.; Miller, J. C. Statistics and Chemometrics for Analytical Chemistry, 5th ed.; Pearson Education Ltd.: Essex, 2005; Chapter 5, pp 115−118.

5645

DOI: 10.1021/acs.analchem.7b00942 Anal. Chem. 2017, 89, 5637−5645