- Email: [email protected]
Highly sensitive detection of organophosphorus pesticides by acetylcholinesterase-coated thin film bulk acoustic resonator mass-loading sensor
Biosensors and Bioelectronics 41 (2013) 163–167
Contents lists available at SciVerse ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Highly sensitive detection of organophosphorus pesticides by acetylcholinesterase-coated thin film bulk acoustic resonator mass-loading sensor Da Chen n, Jingjing Wang, Yan Xu, Dehua Li, Luyin Zhang, Zhaoxin Li Qingdao Key Laboratory of Terahertz Technology, Department of Physics, School of Science, Shandong University of Science and Technology, Qianwangang Road 579 #, Qingdao 266510, PR China
a r t i c l e i n f o
abstract
Article history: Received 8 May 2012 Received in revised form 3 August 2012 Accepted 3 August 2012 Available online 30 August 2012
An acetylcholinesterase-coated thin film bulk acoustic resonator has been developed for the detection of organophosphorus pesticides. The thin film bulk acoustic resonator acts as a robust mass-sensitive transducer for bio-sensing. This device works in thickness shear mode with a resonance at 1.97 GHz. The detection is based on the inhibitory effects of organophosphorus compounds on the enzymatic activity of the acetylcholinesterase immobilized on one of the faces of the acoustic resonator. The enzyme reaction in the substrate solution and the inhibitory effect is observed are real time by measuring the frequency shift. The presence of organophosphorus pesticides can be detected from the diminution of the frequency shift compared with the levels found in their absence. The device exhibits linear responses, good reproducibility, simple operation, portability and a low detection limit of 5.3 10 11 M for paraoxon. The detection results of organophosphorus pesticide residues in practical samples show that the proposed sensor has the feasibility and sensing accuracy comparable to gas chromatography. & 2012 Elsevier B.V. All rights reserved.
Keywords: Organophosphorus pesticides Mass-loading sensors Thin film bulk acoustic resonator Pesticide residue
1. Introduction Organophosphorus (OP) pesticides are a kind of broad-spectrum insecticide used for a wide range of crops, including vegetables, fruits and grains. However, OP pesticides present an acute toxicity that can cause a serious risk to the balance of aquatic systems and even danger to human health through contamination and pesticide residues in farm products. In the past, the standard methods for OP detection included high-pressure liquid chromatography (HPLC) and gas chromatography (GC) in combination with mass spectrometry (GC/MS) (Maurer, 2002; Aprea et al., 2002). Although these methods offer quantitative analysis with sensitivity and selectivity, they are slow, expensive, and laborious. For food safety, quick and accurate tests are essential to allow the detection of contamination in foods before they are distributed to consumers. Therefore, portable biosensors that can be applied for on-site rapid detection are of great practical interest. A number of biosensors with different signal transduction mechanisms, including enzyme-linked immunosorbent assays (ELISA) (Kim et al., 2003; Liu et al., 2009), optics (Soh et al., 2003; Yang et al., 2008), electrochemistry (Mulchandani et al., 2001; Liu and Lin, 2006; Wang et al., 2011), microcantilevers
n
Corresponding author. Tel./fax: þ 86 53286057555. E-mail address: [email protected] (D. Chen).
0956-5663/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.bios.2012.08.018
(Yang et al., 2003; Karnati et al., 2007) and quartz crystal microbalances (QCMs) (Abad et al., 1998; Hu et al., 2005; Kim et al., 2007) have been reported to detect OP compounds. Most of these methods involve the inhibitory effect of OP compounds on the enzymatic activity of acetylcholinesterase (AChE) (Van Dyk and Pletschke, 2011). Among the biochemical detection technologies currently employed, mass-loading sensors have received much attention because they eliminate the use of target labeling and allow realtime monitoring of bio-specific interactions (Ferreira et al., 2009; Arlett et al., 2011). In mass-sensitive biomolecular detection, a specific bio-recognition receptor is immobilized on the surface of a mechanical resonator. As target molecules bind to the coated receptor, the increasing mass causes a decrease in the frequency of the natural mechanical resonance. Using the well-known masssensitive QCMs, the inhibition in the enzymatic activity of AChE with an inhibitor in the presence of a substrate has been measured (Abad et al., 1998). The sensitivity of the QCM based enzyme sensor is amplified significantly by increasing the mass-loading via the precipitation of enzyme reaction products. The detection limits of an AChE-coated QCMs were reported to be 5.0 10 8 and 1.0 10 7 M for paraoxon and carbaryl (Abad et al., 1998), respectively, 1.55 10 7 M for ethyl para-nitro-phenyl (Kim et al., 2007), and 3.63 10 7 M for paraoxon-ethyl (Kim and Kim, 2009). In principle, the sensitivity of mass loading sensors is proportional to a power of the fundamental resonance frequency. However, QCMs
164
D. Chen et al. / Biosensors and Bioelectronics 41 (2013) 163–167
are made from a thin plate of AT-cut quartz and work typically across several megahertz frequencies. It is difficult to further increase the fundamental frequency of the QCMs because of its dependence on the thickness of the AT-cut quartz plate. To overcome this issue, a thin film bulk acoustic resonator (TFBAR) working across the 2–5 GHz range has been developed as an alternative (Gabl et al., 2004; Wingqvist, 2010; Voiculescu and Nordin, 2012). The resonance frequency up to the gigahertz range provides a potential sensitivity increase of more than three orders of magnitude over traditional QCMs, making TFBAR ideal for low concentration biochemical sensing applications. In addition, TFBAR makes use of standard silicon technology, thereby realizing the ability to inexpensively combine a number of sensors on a chip and integrate them with the read-out circuits (Nirschl et al., 2010). In this way, low cost, portable, high throughput and miniature sensor systems can be realized. Today, some attempts to use TFBAR for biochemical gas sensing applications have been made with interesting results (Wingqvist et al., 2007; Nirschl et al., 2009; Wang et al., 2011; Zhang et al., 2010). In our previous paper, we also developed TFBAR sensors to detect trace nerve vapor (Chen et al., 2010), hydrogen (Chen et al., 2011c), and a cancer biomarker (Chen et al., 2011b). These devices were operated in longitudinal mode and consisted of a thin piezoelectric film sandwiched by two electrodes and a Bragg acoustic reflector. The results of previous studies verify that the sensitivity of TFBAR is extremely higher than those of QCMs. In this paper, we report a novel AChE-coated TFBAR mass-loading sensor to detect trace level OP pesticides. Here a shear mode resonator building on the silicon nitride diaphragm was fabricated. Due to the minimal damping of the shear mode wave in the adjacent liquid medium, the shear mode resonator has a relatively higher Qfactor in liquids and is therefore more suitable for real time biological sensing compared with the general longitudinal mode TFBAR. Meanwhile, the detection of OP pesticides is realized by measuring the degree of the enzymatic reaction in real time, based on the inhibitory effect of the OP pesticides on the activity of the AChE immobilized on the sensing surface of TFBAR. Here, the sensing responses, reproducibility, and the accuracy and feasibility of the TFBAR mass-loading sensor in practice were investigated. The high performance of the device shows that our proposed sensor is a promising tool for the rapid and accurate detection of OP pesticides.
2. Experimental 2.1. Reagents A sensing element, comprising acetylcholinesterase (AChE), obtained from an electric eel (Electrophorus electricus), and type VI-S lyophilized powder, were purchased from Sigma-Aldrich (MO, USA). The AChE enzyme used was in a tetrameric form with a molecular weight of 280 kDa. The histological substrate, 3-indolyl acetate (3-IA), 97%, was purchased from Sigma-Aldrich and was used as received. The stock substrate solutions of 3-IA (50 g l 1) were prepared immediately prior to use in dimethylformamide (DMF, Sigma-Aldrich MO, USA). The phosphate buffer solution (PBS) and the reagents for immobilization, including sulfo-succinimidyl-6-[3-(2-pyridyldithio) propionamido] hexanoate (sulfo-LC-SPDP) and dithiothreitol (DTT), were supplied by Pierce (IL, USA). The standard samples of paraoxon for testing were obtained from the Institute of Environmental Protection Agriculture Ministry, China. 2.2. Configuration of the TFBAR mass-loading sensor Fig. 1 shows the basic configuration and a microphotograph of the AChE-coated TFBAR mass-loading sensor. The 1.5-mm-thick
Fig. 1. Basic configuration (a) and a microphotograph (b) of the AChE-coated TFBAR mass-loading sensor. The AChE enzyme is immobilized on the bottom of the testing cavity and exposed to the substrate solution. In the tests, the device was packaged in a PCB connected to a network analyzer and the testing cavity was further connected to a syringe pump to deliver the analyte liquid.
piezoelectric AlN film is built on a 0.8-mm-thick silicon nitride diaphragm deposited by low pressure chemical vapor deposition. The silicon is etched by KOH to isolate the resonator acoustically from the substrate and also create a testing channel above the resonator for the liquid to be analyzed. A 20-nm-thick Au/Ti layer is deposited on the bottom of the channel to immobilize the biomolecules. The thickness shear mode acoustic wave is excited in the AlN film by lateral electric field. The feasibility of this method has been verified by theoretical and experimental analyses reported in our previous paper (Chen et al., 2011a). In this study, the electric field was optimized by finite element modeling according to the method described in another literature (Evgeny et al., 2008). In this experiment, two parallel electrodes were fabricated on the surface of the AlN film with a gap of 10 mm. In the tests, the TFBAR mass-loading sensor was packaged in a PCB connected to a network analyzer (Agilent 8714ET) and the testing cavity was further connected to a syringe pump (Longer LSP01-1A) to deliver the analyte liquid. The frequency response was measured using the network analyzer and the data were loaded to a computer to analyze. 2.3. Sensing principle of the TFBAR mass-loading sensor The sensing principle of the TFBAR mass-loading sensor in this study is based on the inhibitory effect of OP pesticides on the activity of AChE. The AChE is immobilized on the Au surface above the resonator and exposed to the substrate solution containing 3-indoyl acetate. As the enzymatic reaction product (indigo pigment) is precipitated over the resonator surface after dimerization (Abad et al., 1998), the enzymatic reaction can be traced, in real-time and in-situ, by measuring the frequency decrease caused by the increasing mass over the resonator surface. In the presence of OP pesticides, given that they inhibit AChE activity, the detection of AChE inhibitors can be carried out by following the diminution of the frequency shift from the levels found in their absence. The schematic diagram of the sensing mechanism of the AChE-coated TFBAR mass-loading sensor is shown in Supplementary Fig. S1 online. 2.4. Process of enzyme immobilization The immobilization of AChE on the Au surface is based on the chemisorption of AChE enzymes that have been thiolated with a heterobifunctional cross-linker, sulfo-LC-SPDP, according to the method previously described for AChE immobilization on a QCM (Kim et al., 2007). The immobilization process was optimized to obtain a highly active AChE coating by measuring the frequency shift of the TFBAR in the substrate solution as shown in
D. Chen et al. / Biosensors and Bioelectronics 41 (2013) 163–167
Supplementary Fig. S2 online. The sensing Au surface on the bottom of the testing cavity was soaked in 1.2 M NaOH for 5 min, washed with distilled water, and then immersed in 1.2 M HCl for 5 min to clean the surface. After washing with distilled water again, the surface was dried in a convection oven. Then, 20 ml of 50 mg ml 1 AChE PBS solution and the same volume of 20 mM sulfo-LC-SPDP dissolved in distilled water were mixed and the resulting mixture was incubated at room temperature for 1 h. To reduce the disulfide bond of the thiolated AChE, 15 ml of DTT dissolved in 0.1 M sodium acetate buffer (pH 4.5) containing 0.1 M NaCl was added and made to react for 30 min. The resulting solution was spread over the entire testing cavity above the resonator, followed by drying for 1 h at room temperature. After washing with distilled water and PBS, the AChE-coated TFBAR mass-loading sensor was prepared and ready to test. 2.5. Testing procedure of the TFBAR mass-loading sensor Paraoxon, which is one of a number of widely applied OP pesticides for fruit and vegetables, was used as the analyte in this experiment. Before testing, the sensing Au area was immersed in PBS (0.1 M; pH 8) and placed in a thermostated cell at 20 1C. The resonance frequency was monitored for an initial period of 10 min until the signal stabilized for all the measurements. First, the substrate solution (3-IA DMF solution) was injected into the testing cavity to initiate the enzymatic reaction. The resonance frequency was continuously monitored and the time evolution was obtained by plotting the frequency shift versus time. In order to detect the presence of OP pesticides, the mixtures of substrate solution and pesticide with different concentrations (10 10– 10 5 M) were injected. The percent inhibition (%I) was calculated as follows: %I ¼
Df O Df P 100 Df O
where Dfo is the frequency shift obtained in the absence of pesticide and DfP is the frequency shift obtained after the addition of a known concentration of pesticide. 2.6. The OP residue analysis in practice The practical accuracy and feasibility of the TFBAR massloading sensor were evaluated by the detection of residue pesticide in growing radishes. The radishes were first planted in flowerpots in a greenhouse. Twenty days post-sowing, the leaves were sprayed with paraoxon solution (500 mg l 1). The whole plants, including roots and leaves, were sampled for 4, 8 and 12 days after spraying for the analysis of pesticide residue. In order to confirm the accuracy of the TFBAR mass-loading sensor, the same prepared samples were analyzed using a GC instrument (SP-6800A, Ruhong, China). In the experiments of the GC analysis, a flame photometric detector (FPD) and a DB-1701 chromatography column (30 mm 0.32 mm 0.25 mm) were used. The temperature of the chromatographic column was 120 1C at first and was then increased to 275 1C at the heating rate of 10 1C min 1. The FPD temperature was 280 1C. The flow rates of the carrier gas, end-puff, air and hydrogen were 5 ml min 1, 45 ml min 1, 100 ml min 1 and 75 ml min 1, respectively. The process of sample preparation is as follows: first, the surface soil and dirt were removed from the radishes by rinsing in deionized water. The plants were then chopped into pieces of about 1 mm in length and mixed before weighing. The mixed samples 5 g in weight were homogenized in 20 ml ethyl acetate and centrifuged at 3000 rpm for 5 min to produce pellets. The supernatant was transferred into a conical flask and the pellets
165
were extracted one more time with the same amount of ethyl acetate. The supernatants from the first and second extractions were mixed and concentrated to dryness using a rotary evaporator. The dried material in the treatment group was dissolved in 10 ml of deionized water for TFBAR detection, while the dried material in the control group was dissolved in 5 ml of ethyl acetate for GC analysis.
3. Results and discussion 3.1. Performance of the TFBAR mass-loading sensor in air and liquids The TFBAR was fabricated as a robust mass-sensitive transducer for bio-sensing. Fig. 2 shows the conductance curves of the TFBAR without any coating measured in air and in the two solvents used for substrate dissolution (DMF) and the enzymatic reaction (PBS, 0.1 M; pH 8.0). The resonator working in DMF and PBS shows similar conductance responses with the resonance frequency near 1.975 GHz. A frequency shift of 4.5 MHz and a decline of about 10% in the conductance peak are found for the measurements in liquids in comparison with those in air. However, this decay in performance of the proposed shear mode TFBAR is obviously smaller than that of the general longitudinal mode TFBAR (Zhang et al., 2005), which can be attributed to the relatively low damping effect of the liquid on the shear mode acoustic wave. 3.2. Sensor response after AChE immobilization The resonance frequencies of the TFBAR decrease after AChE immobilization on the Au surface caused by the sensitivity change of the mass-loading sensors. The frequency shifts of the TFBAR after AChE immobilization are listed in Table 1, of which the data are calculated based on the average frequency values measured five times for each device. This frequency shift indicates that the AChE enzyme is efficiently bound to the sensing Au surface by the crosslinking method. Furthermore, it is noted that the frequency shifts after AChE coating in air and in the liquids are both about 3 MHz, which suggests that the additions of the above solvents do not cause meaningful change of the frequency shift. From these results, it is believed that the possible interference caused by the addition of solvents might be negligible in this experiment.
Fig. 2. Measured conductance curves of the TFBAR mass-loading sensor without any coating in air, DMF and PBS. The resonator working in DMF and PBS shows similar conductance responses with the resonance frequency near 1.975 GHz.
166
D. Chen et al. / Biosensors and Bioelectronics 41 (2013) 163–167
Table 1 Frequency shifts of the AChE-coated TFBAR mass-loading sensor in the presence of paraoxon. Device
1 2 3 4 5 CV (%)
Frequency shift after AChE coating (MHz)
Frequency shift for different paraoxon concentrations (MHz)
Air
DMF
PBS
Absence
2 10 10
2 10 8
2 10 5
2.92 7 0.07 2.79 7 0.05 2.98 7 0.06 2.76 7 0.10 2.75 7 0.04 3.65
3.04 70.11 2.75 70.06 3.08 70.08 2.94 70.09 2.86 70.06 4.57
2.977 0.04 2.767 0.08 3.047 0.07 2.947 0.10 2.847 0.07 3.79
1.24 7 0.08 1.14 7 0.05 1.27 7 0.04 1.20 7 0.08 1.26 7 0.05 4.34
1.12 7 0.04 1.05 7 0.05 1.18 7 0.07 1.07 7 0.21 1.12 7 0.06 4.58
0.717 0.05 0.697 0.08 0.767 0.10 0.687 0.09 0.687 0.07 4.78
0.25 70.07 0.24 70.04 0.26 70.06 0.27 70.09 0.24 70.06 5.07
3.3. Inhibition study on the TFBAR mass-loading sensor with OP pesticides In the enzymatic reaction, AChE hydrolyzes the ester bond of the substrate, giving rise to acetate and 3-hydroxyindole as by-products. The hydroxyl group of the 3-hydroxyindole product tautomerizes, forming a ketone, which in neutral or alkaline conditions, causes dimerization to occur to form a water insoluble indigo pigment product deposited on the Au surface (Abad et al., 1998; Kim et al., 2007). The formation of the precipitate contributes to an increased loading on the resonator, causing the resonance frequency to decrease during the course of the enzymatic reaction. Fig. 3 shows the time-dependent resonance frequency profiles in the absence and presence of increasing concentrations of paraoxon. The original frequencies in this figure were obtained when the device was immersed in PBS. In the absence of the inhibitor, the frequency decreases rapidly during the course of the reaction and gradually reaches an equilibrium when the reaction time is longer than 40 min; then an approximately stable frequency shift (Dfo ¼1.24 MHz) is obtained. This observation indicates that further enzymatic reaction tends to stop possibly because the sensing Au surface has been completely covered with the precipitate. To rule out the possibility of the frequency shift due to the enzyme substrate solution, the resonance frequency of the bare TFBAR without AChE immobilization was also tested after the injection of the substrate solution. As shown in Fig. 3(e), there is no obvious frequency change for the bare TFBAR in a long time, indicating that the large frequency shift produced on the AChEcoated devices results from the enzyme reaction. In addition, the measurements were performed after several days using the AChEcoated TFBAR devices stored in nitrogen at 4 1C. Under the same experimental conditions, the frequency shift (Dfo) measured after 10 days is 91.2% of that found from the first measurement, indicating that the activity of AChE immobilized on the Au surface remains stable for a long period of time. In order to remove the adsorbed precipitate on the Au surface for further testing, the Au sensing surface was washed with aliquots of DMF several times. After this treatment, the resonance frequency of the resonator increased to the original frequency again, which seemed to indicate that the precipitate on the Au sensing surface was removed by DMF dissolution, as reported previously (Karousos et al., 2002). The inhibitory effects of OP pesticides on AChE activity are well known. Such inhibitory effects can be studied, in-situ and in real time, using our AChE-coated TFBAR mass-loading sensor by following the diminution of frequency shifts derived from the deposited precipitate formed by the enzymatic reaction on the resonator. As shown in Fig. 3, when the substrate solution and paraoxon were injected simultaneously, the change trends of the resonance frequency conspicuously slow down compared with the case of absence of paraoxon. The inhibitory effect of paraoxon on the AChE activity is very obvious.
Fig. 3. Time-dependent resonance frequency profiles in the absence and presence of increasing concentrations of paraoxon: (a) absence; (b) 2 10 10 M; (c) 2 10 8 M and (d) 2 10 5 M. The mixtures of substrate solution and pesticides were injected to the testing cavity simultaneously. The profile (e) is the frequency shift of the bare TFBAR without AChE immobilization after the injection of the substrate solution. Details of the invisible part are shown in Supplementary Fig. S3 online. The inset of this figure is the calibration curve of the percent inhibition (%I) versus paraoxon concentrations. The frequency shifts were determined at 40 min for all the profiles.
To trace the relationship between pesticide concentration and the frequency shift quantitatively, the frequency shifts were determined at 40 min for all the profiles and listed in Table 1. This assay time chosen because the frequency shift in high concentration OP solutions is very small in a short time. Actually, for trace level detections, the frequency shifts can be determined at a shorter time. The inset of Fig. 3 shows the calibration curve of the percent inhibition (%I) versus paraoxon concentrations, of which the error bars are standard deviation (S.D.) of the replicate measurements. A linear relationship on a semi-logarithmic scale was found between the analyte concentration and the percent inhibition, with the regression equation Y ¼13.37X þ142.49 (r ¼0.9954). The limit of detection (LOD), which is usually defined as the concentration of inhibitor required to achieve 5% inhibition (Abad et al., 1998), was safely presumed as 5.3 10 11 M for paraoxon. It is worth noting that this LOD value obtained using our TFBAR massloading sensor is lower by three orders of magnitude than that of the AChE-coated QCMs (Abad et al., 1998; Kim et al., 2007) and comparable to, or lower than, the literature values measured using ELISA, electrochemical and optical methods (Van Dyk and Pletschke, 2011). However, in our case, we are able to employ a significantly smaller sensor size and simpler operation.
D. Chen et al. / Biosensors and Bioelectronics 41 (2013) 163–167
Table 2 Concentration of residual paraoxon on radish samples. Methods
4 days latera ( 10 8 M)
8 days later ( 10 8 M)
12 days later ( 10 9 M)
AChE-coated TFBAR GC Relative deviation (%)
7.27 0.2
1.2 70.1
4.97 0.3
7.97 0.1 8.87
1.1 70.1 9.09
5.37 0.1 7.54
a Days after paraoxon with a concentration of 500 mg l 1 was sprayed on the radish leaves.
3.4. Reproducibility of the TFBAR mass-loading sensor The reproducibility of a biosensor is one of the key parameters for practical applications. Generally, the approach based on inhibition of enzymatic activity shows an inherent high availability for biochemical sensing (Van Dorst et al., 2011). To investigate the reproducibility of the TFBAR mass-loading sensor applied in OP detection, the same samples were determined using five different devices repeatedly under the same conditions. As shown in Table 1, the TFBAR mass-loading sensors offer reliable results for OP detection with coefficients of variability (CV) of 4.58%, 4.78%, and 5.07% corresponding to paraoxon concentrations of 2 10 10, 2 10 8 and 2 10 5 M, respectively. These CV values could be seen to be low enough to show reproducibility, considering the normally accepted 5% level for a reasonable analytical method. 3.5. Comparison of the TFBAR mass-loading sensor with a GC instrument In order to evaluate the accuracy and feasibility of the TFBAR mass-loading sensor in practice, the contrast experiments were performed using our AChE-coated TFBAR mass-loading sensor and a commercial GC instrument. Table 2 shows the measured results of the residual pesticide in radish samples and the relative deviations between these two methods. The tested results of both methods are in reasonable agreement with the relative deviations lower than 10% for the sample collected on the same day, which indicates that the sensing accuracy of the TFBAR mass-loading sensor is as high as that of the GC system. It is feasible to apply our proposed sensor to analyze residual pesticide in practice. Remarkably, considering its cost efficiency, simplicity, and portability, the TFBAR mass-loading sensor shows great promise for the on-site analysis of food safety.
4. Conclusions From this study, we are able to show that the AChE-coated TFBAR mass-loading sensor can be used as a feasible tool for the characterization of pesticide analysis and enzyme reactions. It is verified that the inhibitory effects of OP compounds on AChE activity can be studied by following the resonance frequency of the TFBAR during the enzymatic reaction. The TFBAR mass-loading sensor exhibits a remarkably low detection limit, a linear response, good reproducibility, and accuracy comparable to that of the GC method in practice. Significantly, we are able to employ small sensor size, simple operation and low cost to realize in-situ and in real-time detection. Based on the universality of mass sensitivity, this method could possibly be extended to detect a wide variety of biological reactions, such as antigen–antibody binding, protein–ligand interactions and genetic hybridizing, which could provide information for the studies of biological reaction kinetics. The limitation of this study is that we used the AChE enzyme as the sensitive coating of
167
the TFBAR, which cannot identify the OP and carbamates pesticides. In addition, it may not be so convenient to use AChE-coated biosensors for the practical on-site analysis in complex environment because they require a previous comparison with the native enzyme activity. In order to overcome these issues, the antibodies will be employed as the sensitive coating in our further studies.
Acknowledgments This work was supported by National Science Foundation of China (No. 61101027), Qingdao Science and Technology Program of Basic Research Projects [11-2-4-4-(6)-jch and 12-1-4-6-(8)-jch], the Shangdong Province Young and Middle-Aged Scientists Research Awards Fund (BS2012DX032) and SDUST Research Fund (2011KYJQ101).
Appendix A. Supplementary materials Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2012.08.018.
References ˜ a, H.D., Lorenzo, E., 1998. Analytical Abad, J.M., Pariente, F., Herna´ndez, L., Abrun Chemistry 70, 2848–2855. Aprea, C., Colosio, C., Mammone, T., Minoia, C., Maroni, M., 2002. Journal of Chromatography B 769, 191–219. Arlett, J.L., Myers, E.B., Roukes, M.L., 2011. Nature Nanotechnology 6, 203–215. Chen, D., Wang, J., Li, D., Xu, Y., Li, Z., 2011a. Sensors and Actuators A 165, 379–384. Chen, D., Wang, J.J., Li, D.H., Li, Z.X., 2011b. Electronics Letters 47, 1169 -U1123. Chen, D., Wang, J.J., Liu, Q.X., Xu, Y., Li, D.H., Liu, Y.J., 2011c. Journal of Micromechanics and Microengineering 21, 1150181–1150187. Chen, D., Xu, Y., Wang, J., Zhang, L., 2010. Sensors and Actuators B 150, 483–486. Evgeny, M., Sandrine, G., Paul, M., 2008. Journal of Applied Physics 104, 084508–084515. Ferreira, G.N.M., da-Silva, A.C., Tome´, B., 2009. Trends Biotechnology 27, 689–697. Gabl, R., Feucht, H.-D., Zeininger, H., Eckstein, G., Schreiter, M., Primig, R., Pitzer, D., Wersing, W., 2004. Biosensors and Bioelectronics 19, 615–620. Hu, Y., Pinkham, W., French Jr, L.A., Frankel, D., Vetelino, J.F., 2005. Sensors and Actuators B 108, 910–916. Karnati, C., Du, H., Ji, H.F., Xu, X., Lvov, Y., Mulchandani, A., Mulchandani, P., Chen, W., 2007. Biosensors and Bioelectronics 22, 2636–2642. Karousos, N.G., Aouabdi, S., Way, A.S., Reddy, S.M., 2002. Analytica Chimica Acta 469, 189–196. Kim, D.K., Kim, N., 2009. BioChip Journal 2, 218–221. Kim, N., Park, I.S., Kim, D.K., 2007. Biosensors and Bioelectronics 22, 1593–1599. Kim, Y.J., Cho, Y.A., Lee, H.S., Lee, Y.T., 2003. Analytica Chimica Acta 494, 29–40. Liu, G., Lin, Y., 2006. Analytical Chemistry 78, 835–843. Liu, Y., Wang, H., Huang, J., Yang, J., Liu, B., Yang, P., 2009. Analytica Chimica Acta 650, 77–82. Maurer, H.H., 2002. Therapeutic Drug Monitoring 24, 247–254. Mulchandani, A., Chen, W., Mulchandani, P., Wang, J., Rogers, K.R., 2001. Biosensors and Bioelectronics 16, 225–230. Nirschl, M., Blleer, A., Erler, C., Katzschner, B., Vikholm-Lundin, I., Auer, S., Vertwl, J., Pompe, W., Schreiter, M., Mertig, M., 2009. Sensors and Actuators A 156, 180–184. Nirschl, M., Rantala, A., Tukkiniemi, K., Auer, S., Hellgren, A.C., Pitzer, D., Schreiter, M., Vikholm-Lundin, I., 2010. Sensors 10, 4180–4193. Soh, N., Tokuda, T., Watanabe, T., Mishima, K., Imato, T., Masadome, T., Asano, Y., Okutani, S., Niwa, O., Brown, S., 2003. Talanta 60, 733–745. Van Dorst, B., Mehta, J., Bekaert, K., Rouah-Martin, E., De Coen, W., Dubruel, P., Blust, R., Robbens, J., 2011. Biosensors and Bioelectronics 26, 1178–1194. Van Dyk, J.S., Pletschke, B., 2011. Chemosphere 82, 291–307. Voiculescu, I., Nordin, A.N., 2012. Biosensors and Bioelectronics 33, 1–9. Wang, J., Chen, D., Gan, Y., Sun, X., Jin, Y., 2011. Applied Surface Science 257, 4365–4369. Wingqvist, G., 2010. Surface and Coatings Technology 205, 1279–1286. ¨ Wingqvist, G., Bjurstroom, J., Liljeholm, L., Yantchev, V., Katardjiev, I., 2007. Sensors and Actuators B 123, 466–473. Yang, G., White, I.M., Fan, X., 2008. Sensors and Actuators B 133, 105–112. Yang, Y., Ji, H.F., Thundat, T., 2003. Journal of American Chemical Society 125, 1124–1125. Zhang, H., Marma, M.S., Kim, E.S., McKenna, C.E., Thompson, M.E., 2005. Journal of Micromechanics and Microengineering 15, 1911–1916. Zhang, H., Pang, W., Marma, M.S., Lee, C.-Y., Kamal-Bahl, S., Kim, E.S., McKenna, C.E., 2010. Applied Physics Letters 96, 1237021–1237025.
Contents lists available at SciVerse ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Highly sensitive detection of organophosphorus pesticides by acetylcholinesterase-coated thin film bulk acoustic resonator mass-loading sensor Da Chen n, Jingjing Wang, Yan Xu, Dehua Li, Luyin Zhang, Zhaoxin Li Qingdao Key Laboratory of Terahertz Technology, Department of Physics, School of Science, Shandong University of Science and Technology, Qianwangang Road 579 #, Qingdao 266510, PR China
a r t i c l e i n f o
abstract
Article history: Received 8 May 2012 Received in revised form 3 August 2012 Accepted 3 August 2012 Available online 30 August 2012
An acetylcholinesterase-coated thin film bulk acoustic resonator has been developed for the detection of organophosphorus pesticides. The thin film bulk acoustic resonator acts as a robust mass-sensitive transducer for bio-sensing. This device works in thickness shear mode with a resonance at 1.97 GHz. The detection is based on the inhibitory effects of organophosphorus compounds on the enzymatic activity of the acetylcholinesterase immobilized on one of the faces of the acoustic resonator. The enzyme reaction in the substrate solution and the inhibitory effect is observed are real time by measuring the frequency shift. The presence of organophosphorus pesticides can be detected from the diminution of the frequency shift compared with the levels found in their absence. The device exhibits linear responses, good reproducibility, simple operation, portability and a low detection limit of 5.3 10 11 M for paraoxon. The detection results of organophosphorus pesticide residues in practical samples show that the proposed sensor has the feasibility and sensing accuracy comparable to gas chromatography. & 2012 Elsevier B.V. All rights reserved.
Keywords: Organophosphorus pesticides Mass-loading sensors Thin film bulk acoustic resonator Pesticide residue
1. Introduction Organophosphorus (OP) pesticides are a kind of broad-spectrum insecticide used for a wide range of crops, including vegetables, fruits and grains. However, OP pesticides present an acute toxicity that can cause a serious risk to the balance of aquatic systems and even danger to human health through contamination and pesticide residues in farm products. In the past, the standard methods for OP detection included high-pressure liquid chromatography (HPLC) and gas chromatography (GC) in combination with mass spectrometry (GC/MS) (Maurer, 2002; Aprea et al., 2002). Although these methods offer quantitative analysis with sensitivity and selectivity, they are slow, expensive, and laborious. For food safety, quick and accurate tests are essential to allow the detection of contamination in foods before they are distributed to consumers. Therefore, portable biosensors that can be applied for on-site rapid detection are of great practical interest. A number of biosensors with different signal transduction mechanisms, including enzyme-linked immunosorbent assays (ELISA) (Kim et al., 2003; Liu et al., 2009), optics (Soh et al., 2003; Yang et al., 2008), electrochemistry (Mulchandani et al., 2001; Liu and Lin, 2006; Wang et al., 2011), microcantilevers
n
Corresponding author. Tel./fax: þ 86 53286057555. E-mail address: [email protected] (D. Chen).
0956-5663/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.bios.2012.08.018
(Yang et al., 2003; Karnati et al., 2007) and quartz crystal microbalances (QCMs) (Abad et al., 1998; Hu et al., 2005; Kim et al., 2007) have been reported to detect OP compounds. Most of these methods involve the inhibitory effect of OP compounds on the enzymatic activity of acetylcholinesterase (AChE) (Van Dyk and Pletschke, 2011). Among the biochemical detection technologies currently employed, mass-loading sensors have received much attention because they eliminate the use of target labeling and allow realtime monitoring of bio-specific interactions (Ferreira et al., 2009; Arlett et al., 2011). In mass-sensitive biomolecular detection, a specific bio-recognition receptor is immobilized on the surface of a mechanical resonator. As target molecules bind to the coated receptor, the increasing mass causes a decrease in the frequency of the natural mechanical resonance. Using the well-known masssensitive QCMs, the inhibition in the enzymatic activity of AChE with an inhibitor in the presence of a substrate has been measured (Abad et al., 1998). The sensitivity of the QCM based enzyme sensor is amplified significantly by increasing the mass-loading via the precipitation of enzyme reaction products. The detection limits of an AChE-coated QCMs were reported to be 5.0 10 8 and 1.0 10 7 M for paraoxon and carbaryl (Abad et al., 1998), respectively, 1.55 10 7 M for ethyl para-nitro-phenyl (Kim et al., 2007), and 3.63 10 7 M for paraoxon-ethyl (Kim and Kim, 2009). In principle, the sensitivity of mass loading sensors is proportional to a power of the fundamental resonance frequency. However, QCMs
164
D. Chen et al. / Biosensors and Bioelectronics 41 (2013) 163–167
are made from a thin plate of AT-cut quartz and work typically across several megahertz frequencies. It is difficult to further increase the fundamental frequency of the QCMs because of its dependence on the thickness of the AT-cut quartz plate. To overcome this issue, a thin film bulk acoustic resonator (TFBAR) working across the 2–5 GHz range has been developed as an alternative (Gabl et al., 2004; Wingqvist, 2010; Voiculescu and Nordin, 2012). The resonance frequency up to the gigahertz range provides a potential sensitivity increase of more than three orders of magnitude over traditional QCMs, making TFBAR ideal for low concentration biochemical sensing applications. In addition, TFBAR makes use of standard silicon technology, thereby realizing the ability to inexpensively combine a number of sensors on a chip and integrate them with the read-out circuits (Nirschl et al., 2010). In this way, low cost, portable, high throughput and miniature sensor systems can be realized. Today, some attempts to use TFBAR for biochemical gas sensing applications have been made with interesting results (Wingqvist et al., 2007; Nirschl et al., 2009; Wang et al., 2011; Zhang et al., 2010). In our previous paper, we also developed TFBAR sensors to detect trace nerve vapor (Chen et al., 2010), hydrogen (Chen et al., 2011c), and a cancer biomarker (Chen et al., 2011b). These devices were operated in longitudinal mode and consisted of a thin piezoelectric film sandwiched by two electrodes and a Bragg acoustic reflector. The results of previous studies verify that the sensitivity of TFBAR is extremely higher than those of QCMs. In this paper, we report a novel AChE-coated TFBAR mass-loading sensor to detect trace level OP pesticides. Here a shear mode resonator building on the silicon nitride diaphragm was fabricated. Due to the minimal damping of the shear mode wave in the adjacent liquid medium, the shear mode resonator has a relatively higher Qfactor in liquids and is therefore more suitable for real time biological sensing compared with the general longitudinal mode TFBAR. Meanwhile, the detection of OP pesticides is realized by measuring the degree of the enzymatic reaction in real time, based on the inhibitory effect of the OP pesticides on the activity of the AChE immobilized on the sensing surface of TFBAR. Here, the sensing responses, reproducibility, and the accuracy and feasibility of the TFBAR mass-loading sensor in practice were investigated. The high performance of the device shows that our proposed sensor is a promising tool for the rapid and accurate detection of OP pesticides.
2. Experimental 2.1. Reagents A sensing element, comprising acetylcholinesterase (AChE), obtained from an electric eel (Electrophorus electricus), and type VI-S lyophilized powder, were purchased from Sigma-Aldrich (MO, USA). The AChE enzyme used was in a tetrameric form with a molecular weight of 280 kDa. The histological substrate, 3-indolyl acetate (3-IA), 97%, was purchased from Sigma-Aldrich and was used as received. The stock substrate solutions of 3-IA (50 g l 1) were prepared immediately prior to use in dimethylformamide (DMF, Sigma-Aldrich MO, USA). The phosphate buffer solution (PBS) and the reagents for immobilization, including sulfo-succinimidyl-6-[3-(2-pyridyldithio) propionamido] hexanoate (sulfo-LC-SPDP) and dithiothreitol (DTT), were supplied by Pierce (IL, USA). The standard samples of paraoxon for testing were obtained from the Institute of Environmental Protection Agriculture Ministry, China. 2.2. Configuration of the TFBAR mass-loading sensor Fig. 1 shows the basic configuration and a microphotograph of the AChE-coated TFBAR mass-loading sensor. The 1.5-mm-thick
Fig. 1. Basic configuration (a) and a microphotograph (b) of the AChE-coated TFBAR mass-loading sensor. The AChE enzyme is immobilized on the bottom of the testing cavity and exposed to the substrate solution. In the tests, the device was packaged in a PCB connected to a network analyzer and the testing cavity was further connected to a syringe pump to deliver the analyte liquid.
piezoelectric AlN film is built on a 0.8-mm-thick silicon nitride diaphragm deposited by low pressure chemical vapor deposition. The silicon is etched by KOH to isolate the resonator acoustically from the substrate and also create a testing channel above the resonator for the liquid to be analyzed. A 20-nm-thick Au/Ti layer is deposited on the bottom of the channel to immobilize the biomolecules. The thickness shear mode acoustic wave is excited in the AlN film by lateral electric field. The feasibility of this method has been verified by theoretical and experimental analyses reported in our previous paper (Chen et al., 2011a). In this study, the electric field was optimized by finite element modeling according to the method described in another literature (Evgeny et al., 2008). In this experiment, two parallel electrodes were fabricated on the surface of the AlN film with a gap of 10 mm. In the tests, the TFBAR mass-loading sensor was packaged in a PCB connected to a network analyzer (Agilent 8714ET) and the testing cavity was further connected to a syringe pump (Longer LSP01-1A) to deliver the analyte liquid. The frequency response was measured using the network analyzer and the data were loaded to a computer to analyze. 2.3. Sensing principle of the TFBAR mass-loading sensor The sensing principle of the TFBAR mass-loading sensor in this study is based on the inhibitory effect of OP pesticides on the activity of AChE. The AChE is immobilized on the Au surface above the resonator and exposed to the substrate solution containing 3-indoyl acetate. As the enzymatic reaction product (indigo pigment) is precipitated over the resonator surface after dimerization (Abad et al., 1998), the enzymatic reaction can be traced, in real-time and in-situ, by measuring the frequency decrease caused by the increasing mass over the resonator surface. In the presence of OP pesticides, given that they inhibit AChE activity, the detection of AChE inhibitors can be carried out by following the diminution of the frequency shift from the levels found in their absence. The schematic diagram of the sensing mechanism of the AChE-coated TFBAR mass-loading sensor is shown in Supplementary Fig. S1 online. 2.4. Process of enzyme immobilization The immobilization of AChE on the Au surface is based on the chemisorption of AChE enzymes that have been thiolated with a heterobifunctional cross-linker, sulfo-LC-SPDP, according to the method previously described for AChE immobilization on a QCM (Kim et al., 2007). The immobilization process was optimized to obtain a highly active AChE coating by measuring the frequency shift of the TFBAR in the substrate solution as shown in
D. Chen et al. / Biosensors and Bioelectronics 41 (2013) 163–167
Supplementary Fig. S2 online. The sensing Au surface on the bottom of the testing cavity was soaked in 1.2 M NaOH for 5 min, washed with distilled water, and then immersed in 1.2 M HCl for 5 min to clean the surface. After washing with distilled water again, the surface was dried in a convection oven. Then, 20 ml of 50 mg ml 1 AChE PBS solution and the same volume of 20 mM sulfo-LC-SPDP dissolved in distilled water were mixed and the resulting mixture was incubated at room temperature for 1 h. To reduce the disulfide bond of the thiolated AChE, 15 ml of DTT dissolved in 0.1 M sodium acetate buffer (pH 4.5) containing 0.1 M NaCl was added and made to react for 30 min. The resulting solution was spread over the entire testing cavity above the resonator, followed by drying for 1 h at room temperature. After washing with distilled water and PBS, the AChE-coated TFBAR mass-loading sensor was prepared and ready to test. 2.5. Testing procedure of the TFBAR mass-loading sensor Paraoxon, which is one of a number of widely applied OP pesticides for fruit and vegetables, was used as the analyte in this experiment. Before testing, the sensing Au area was immersed in PBS (0.1 M; pH 8) and placed in a thermostated cell at 20 1C. The resonance frequency was monitored for an initial period of 10 min until the signal stabilized for all the measurements. First, the substrate solution (3-IA DMF solution) was injected into the testing cavity to initiate the enzymatic reaction. The resonance frequency was continuously monitored and the time evolution was obtained by plotting the frequency shift versus time. In order to detect the presence of OP pesticides, the mixtures of substrate solution and pesticide with different concentrations (10 10– 10 5 M) were injected. The percent inhibition (%I) was calculated as follows: %I ¼
Df O Df P 100 Df O
where Dfo is the frequency shift obtained in the absence of pesticide and DfP is the frequency shift obtained after the addition of a known concentration of pesticide. 2.6. The OP residue analysis in practice The practical accuracy and feasibility of the TFBAR massloading sensor were evaluated by the detection of residue pesticide in growing radishes. The radishes were first planted in flowerpots in a greenhouse. Twenty days post-sowing, the leaves were sprayed with paraoxon solution (500 mg l 1). The whole plants, including roots and leaves, were sampled for 4, 8 and 12 days after spraying for the analysis of pesticide residue. In order to confirm the accuracy of the TFBAR mass-loading sensor, the same prepared samples were analyzed using a GC instrument (SP-6800A, Ruhong, China). In the experiments of the GC analysis, a flame photometric detector (FPD) and a DB-1701 chromatography column (30 mm 0.32 mm 0.25 mm) were used. The temperature of the chromatographic column was 120 1C at first and was then increased to 275 1C at the heating rate of 10 1C min 1. The FPD temperature was 280 1C. The flow rates of the carrier gas, end-puff, air and hydrogen were 5 ml min 1, 45 ml min 1, 100 ml min 1 and 75 ml min 1, respectively. The process of sample preparation is as follows: first, the surface soil and dirt were removed from the radishes by rinsing in deionized water. The plants were then chopped into pieces of about 1 mm in length and mixed before weighing. The mixed samples 5 g in weight were homogenized in 20 ml ethyl acetate and centrifuged at 3000 rpm for 5 min to produce pellets. The supernatant was transferred into a conical flask and the pellets
165
were extracted one more time with the same amount of ethyl acetate. The supernatants from the first and second extractions were mixed and concentrated to dryness using a rotary evaporator. The dried material in the treatment group was dissolved in 10 ml of deionized water for TFBAR detection, while the dried material in the control group was dissolved in 5 ml of ethyl acetate for GC analysis.
3. Results and discussion 3.1. Performance of the TFBAR mass-loading sensor in air and liquids The TFBAR was fabricated as a robust mass-sensitive transducer for bio-sensing. Fig. 2 shows the conductance curves of the TFBAR without any coating measured in air and in the two solvents used for substrate dissolution (DMF) and the enzymatic reaction (PBS, 0.1 M; pH 8.0). The resonator working in DMF and PBS shows similar conductance responses with the resonance frequency near 1.975 GHz. A frequency shift of 4.5 MHz and a decline of about 10% in the conductance peak are found for the measurements in liquids in comparison with those in air. However, this decay in performance of the proposed shear mode TFBAR is obviously smaller than that of the general longitudinal mode TFBAR (Zhang et al., 2005), which can be attributed to the relatively low damping effect of the liquid on the shear mode acoustic wave. 3.2. Sensor response after AChE immobilization The resonance frequencies of the TFBAR decrease after AChE immobilization on the Au surface caused by the sensitivity change of the mass-loading sensors. The frequency shifts of the TFBAR after AChE immobilization are listed in Table 1, of which the data are calculated based on the average frequency values measured five times for each device. This frequency shift indicates that the AChE enzyme is efficiently bound to the sensing Au surface by the crosslinking method. Furthermore, it is noted that the frequency shifts after AChE coating in air and in the liquids are both about 3 MHz, which suggests that the additions of the above solvents do not cause meaningful change of the frequency shift. From these results, it is believed that the possible interference caused by the addition of solvents might be negligible in this experiment.
Fig. 2. Measured conductance curves of the TFBAR mass-loading sensor without any coating in air, DMF and PBS. The resonator working in DMF and PBS shows similar conductance responses with the resonance frequency near 1.975 GHz.
166
D. Chen et al. / Biosensors and Bioelectronics 41 (2013) 163–167
Table 1 Frequency shifts of the AChE-coated TFBAR mass-loading sensor in the presence of paraoxon. Device
1 2 3 4 5 CV (%)
Frequency shift after AChE coating (MHz)
Frequency shift for different paraoxon concentrations (MHz)
Air
DMF
PBS
Absence
2 10 10
2 10 8
2 10 5
2.92 7 0.07 2.79 7 0.05 2.98 7 0.06 2.76 7 0.10 2.75 7 0.04 3.65
3.04 70.11 2.75 70.06 3.08 70.08 2.94 70.09 2.86 70.06 4.57
2.977 0.04 2.767 0.08 3.047 0.07 2.947 0.10 2.847 0.07 3.79
1.24 7 0.08 1.14 7 0.05 1.27 7 0.04 1.20 7 0.08 1.26 7 0.05 4.34
1.12 7 0.04 1.05 7 0.05 1.18 7 0.07 1.07 7 0.21 1.12 7 0.06 4.58
0.717 0.05 0.697 0.08 0.767 0.10 0.687 0.09 0.687 0.07 4.78
0.25 70.07 0.24 70.04 0.26 70.06 0.27 70.09 0.24 70.06 5.07
3.3. Inhibition study on the TFBAR mass-loading sensor with OP pesticides In the enzymatic reaction, AChE hydrolyzes the ester bond of the substrate, giving rise to acetate and 3-hydroxyindole as by-products. The hydroxyl group of the 3-hydroxyindole product tautomerizes, forming a ketone, which in neutral or alkaline conditions, causes dimerization to occur to form a water insoluble indigo pigment product deposited on the Au surface (Abad et al., 1998; Kim et al., 2007). The formation of the precipitate contributes to an increased loading on the resonator, causing the resonance frequency to decrease during the course of the enzymatic reaction. Fig. 3 shows the time-dependent resonance frequency profiles in the absence and presence of increasing concentrations of paraoxon. The original frequencies in this figure were obtained when the device was immersed in PBS. In the absence of the inhibitor, the frequency decreases rapidly during the course of the reaction and gradually reaches an equilibrium when the reaction time is longer than 40 min; then an approximately stable frequency shift (Dfo ¼1.24 MHz) is obtained. This observation indicates that further enzymatic reaction tends to stop possibly because the sensing Au surface has been completely covered with the precipitate. To rule out the possibility of the frequency shift due to the enzyme substrate solution, the resonance frequency of the bare TFBAR without AChE immobilization was also tested after the injection of the substrate solution. As shown in Fig. 3(e), there is no obvious frequency change for the bare TFBAR in a long time, indicating that the large frequency shift produced on the AChEcoated devices results from the enzyme reaction. In addition, the measurements were performed after several days using the AChEcoated TFBAR devices stored in nitrogen at 4 1C. Under the same experimental conditions, the frequency shift (Dfo) measured after 10 days is 91.2% of that found from the first measurement, indicating that the activity of AChE immobilized on the Au surface remains stable for a long period of time. In order to remove the adsorbed precipitate on the Au surface for further testing, the Au sensing surface was washed with aliquots of DMF several times. After this treatment, the resonance frequency of the resonator increased to the original frequency again, which seemed to indicate that the precipitate on the Au sensing surface was removed by DMF dissolution, as reported previously (Karousos et al., 2002). The inhibitory effects of OP pesticides on AChE activity are well known. Such inhibitory effects can be studied, in-situ and in real time, using our AChE-coated TFBAR mass-loading sensor by following the diminution of frequency shifts derived from the deposited precipitate formed by the enzymatic reaction on the resonator. As shown in Fig. 3, when the substrate solution and paraoxon were injected simultaneously, the change trends of the resonance frequency conspicuously slow down compared with the case of absence of paraoxon. The inhibitory effect of paraoxon on the AChE activity is very obvious.
Fig. 3. Time-dependent resonance frequency profiles in the absence and presence of increasing concentrations of paraoxon: (a) absence; (b) 2 10 10 M; (c) 2 10 8 M and (d) 2 10 5 M. The mixtures of substrate solution and pesticides were injected to the testing cavity simultaneously. The profile (e) is the frequency shift of the bare TFBAR without AChE immobilization after the injection of the substrate solution. Details of the invisible part are shown in Supplementary Fig. S3 online. The inset of this figure is the calibration curve of the percent inhibition (%I) versus paraoxon concentrations. The frequency shifts were determined at 40 min for all the profiles.
To trace the relationship between pesticide concentration and the frequency shift quantitatively, the frequency shifts were determined at 40 min for all the profiles and listed in Table 1. This assay time chosen because the frequency shift in high concentration OP solutions is very small in a short time. Actually, for trace level detections, the frequency shifts can be determined at a shorter time. The inset of Fig. 3 shows the calibration curve of the percent inhibition (%I) versus paraoxon concentrations, of which the error bars are standard deviation (S.D.) of the replicate measurements. A linear relationship on a semi-logarithmic scale was found between the analyte concentration and the percent inhibition, with the regression equation Y ¼13.37X þ142.49 (r ¼0.9954). The limit of detection (LOD), which is usually defined as the concentration of inhibitor required to achieve 5% inhibition (Abad et al., 1998), was safely presumed as 5.3 10 11 M for paraoxon. It is worth noting that this LOD value obtained using our TFBAR massloading sensor is lower by three orders of magnitude than that of the AChE-coated QCMs (Abad et al., 1998; Kim et al., 2007) and comparable to, or lower than, the literature values measured using ELISA, electrochemical and optical methods (Van Dyk and Pletschke, 2011). However, in our case, we are able to employ a significantly smaller sensor size and simpler operation.
D. Chen et al. / Biosensors and Bioelectronics 41 (2013) 163–167
Table 2 Concentration of residual paraoxon on radish samples. Methods
4 days latera ( 10 8 M)
8 days later ( 10 8 M)
12 days later ( 10 9 M)
AChE-coated TFBAR GC Relative deviation (%)
7.27 0.2
1.2 70.1
4.97 0.3
7.97 0.1 8.87
1.1 70.1 9.09
5.37 0.1 7.54
a Days after paraoxon with a concentration of 500 mg l 1 was sprayed on the radish leaves.
3.4. Reproducibility of the TFBAR mass-loading sensor The reproducibility of a biosensor is one of the key parameters for practical applications. Generally, the approach based on inhibition of enzymatic activity shows an inherent high availability for biochemical sensing (Van Dorst et al., 2011). To investigate the reproducibility of the TFBAR mass-loading sensor applied in OP detection, the same samples were determined using five different devices repeatedly under the same conditions. As shown in Table 1, the TFBAR mass-loading sensors offer reliable results for OP detection with coefficients of variability (CV) of 4.58%, 4.78%, and 5.07% corresponding to paraoxon concentrations of 2 10 10, 2 10 8 and 2 10 5 M, respectively. These CV values could be seen to be low enough to show reproducibility, considering the normally accepted 5% level for a reasonable analytical method. 3.5. Comparison of the TFBAR mass-loading sensor with a GC instrument In order to evaluate the accuracy and feasibility of the TFBAR mass-loading sensor in practice, the contrast experiments were performed using our AChE-coated TFBAR mass-loading sensor and a commercial GC instrument. Table 2 shows the measured results of the residual pesticide in radish samples and the relative deviations between these two methods. The tested results of both methods are in reasonable agreement with the relative deviations lower than 10% for the sample collected on the same day, which indicates that the sensing accuracy of the TFBAR mass-loading sensor is as high as that of the GC system. It is feasible to apply our proposed sensor to analyze residual pesticide in practice. Remarkably, considering its cost efficiency, simplicity, and portability, the TFBAR mass-loading sensor shows great promise for the on-site analysis of food safety.
4. Conclusions From this study, we are able to show that the AChE-coated TFBAR mass-loading sensor can be used as a feasible tool for the characterization of pesticide analysis and enzyme reactions. It is verified that the inhibitory effects of OP compounds on AChE activity can be studied by following the resonance frequency of the TFBAR during the enzymatic reaction. The TFBAR mass-loading sensor exhibits a remarkably low detection limit, a linear response, good reproducibility, and accuracy comparable to that of the GC method in practice. Significantly, we are able to employ small sensor size, simple operation and low cost to realize in-situ and in real-time detection. Based on the universality of mass sensitivity, this method could possibly be extended to detect a wide variety of biological reactions, such as antigen–antibody binding, protein–ligand interactions and genetic hybridizing, which could provide information for the studies of biological reaction kinetics. The limitation of this study is that we used the AChE enzyme as the sensitive coating of
167
the TFBAR, which cannot identify the OP and carbamates pesticides. In addition, it may not be so convenient to use AChE-coated biosensors for the practical on-site analysis in complex environment because they require a previous comparison with the native enzyme activity. In order to overcome these issues, the antibodies will be employed as the sensitive coating in our further studies.
Acknowledgments This work was supported by National Science Foundation of China (No. 61101027), Qingdao Science and Technology Program of Basic Research Projects [11-2-4-4-(6)-jch and 12-1-4-6-(8)-jch], the Shangdong Province Young and Middle-Aged Scientists Research Awards Fund (BS2012DX032) and SDUST Research Fund (2011KYJQ101).
Appendix A. Supplementary materials Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2012.08.018.
References ˜ a, H.D., Lorenzo, E., 1998. Analytical Abad, J.M., Pariente, F., Herna´ndez, L., Abrun Chemistry 70, 2848–2855. Aprea, C., Colosio, C., Mammone, T., Minoia, C., Maroni, M., 2002. Journal of Chromatography B 769, 191–219. Arlett, J.L., Myers, E.B., Roukes, M.L., 2011. Nature Nanotechnology 6, 203–215. Chen, D., Wang, J., Li, D., Xu, Y., Li, Z., 2011a. Sensors and Actuators A 165, 379–384. Chen, D., Wang, J.J., Li, D.H., Li, Z.X., 2011b. Electronics Letters 47, 1169 -U1123. Chen, D., Wang, J.J., Liu, Q.X., Xu, Y., Li, D.H., Liu, Y.J., 2011c. Journal of Micromechanics and Microengineering 21, 1150181–1150187. Chen, D., Xu, Y., Wang, J., Zhang, L., 2010. Sensors and Actuators B 150, 483–486. Evgeny, M., Sandrine, G., Paul, M., 2008. Journal of Applied Physics 104, 084508–084515. Ferreira, G.N.M., da-Silva, A.C., Tome´, B., 2009. Trends Biotechnology 27, 689–697. Gabl, R., Feucht, H.-D., Zeininger, H., Eckstein, G., Schreiter, M., Primig, R., Pitzer, D., Wersing, W., 2004. Biosensors and Bioelectronics 19, 615–620. Hu, Y., Pinkham, W., French Jr, L.A., Frankel, D., Vetelino, J.F., 2005. Sensors and Actuators B 108, 910–916. Karnati, C., Du, H., Ji, H.F., Xu, X., Lvov, Y., Mulchandani, A., Mulchandani, P., Chen, W., 2007. Biosensors and Bioelectronics 22, 2636–2642. Karousos, N.G., Aouabdi, S., Way, A.S., Reddy, S.M., 2002. Analytica Chimica Acta 469, 189–196. Kim, D.K., Kim, N., 2009. BioChip Journal 2, 218–221. Kim, N., Park, I.S., Kim, D.K., 2007. Biosensors and Bioelectronics 22, 1593–1599. Kim, Y.J., Cho, Y.A., Lee, H.S., Lee, Y.T., 2003. Analytica Chimica Acta 494, 29–40. Liu, G., Lin, Y., 2006. Analytical Chemistry 78, 835–843. Liu, Y., Wang, H., Huang, J., Yang, J., Liu, B., Yang, P., 2009. Analytica Chimica Acta 650, 77–82. Maurer, H.H., 2002. Therapeutic Drug Monitoring 24, 247–254. Mulchandani, A., Chen, W., Mulchandani, P., Wang, J., Rogers, K.R., 2001. Biosensors and Bioelectronics 16, 225–230. Nirschl, M., Blleer, A., Erler, C., Katzschner, B., Vikholm-Lundin, I., Auer, S., Vertwl, J., Pompe, W., Schreiter, M., Mertig, M., 2009. Sensors and Actuators A 156, 180–184. Nirschl, M., Rantala, A., Tukkiniemi, K., Auer, S., Hellgren, A.C., Pitzer, D., Schreiter, M., Vikholm-Lundin, I., 2010. Sensors 10, 4180–4193. Soh, N., Tokuda, T., Watanabe, T., Mishima, K., Imato, T., Masadome, T., Asano, Y., Okutani, S., Niwa, O., Brown, S., 2003. Talanta 60, 733–745. Van Dorst, B., Mehta, J., Bekaert, K., Rouah-Martin, E., De Coen, W., Dubruel, P., Blust, R., Robbens, J., 2011. Biosensors and Bioelectronics 26, 1178–1194. Van Dyk, J.S., Pletschke, B., 2011. Chemosphere 82, 291–307. Voiculescu, I., Nordin, A.N., 2012. Biosensors and Bioelectronics 33, 1–9. Wang, J., Chen, D., Gan, Y., Sun, X., Jin, Y., 2011. Applied Surface Science 257, 4365–4369. Wingqvist, G., 2010. Surface and Coatings Technology 205, 1279–1286. ¨ Wingqvist, G., Bjurstroom, J., Liljeholm, L., Yantchev, V., Katardjiev, I., 2007. Sensors and Actuators B 123, 466–473. Yang, G., White, I.M., Fan, X., 2008. Sensors and Actuators B 133, 105–112. Yang, Y., Ji, H.F., Thundat, T., 2003. Journal of American Chemical Society 125, 1124–1125. Zhang, H., Marma, M.S., Kim, E.S., McKenna, C.E., Thompson, M.E., 2005. Journal of Micromechanics and Microengineering 15, 1911–1916. Zhang, H., Pang, W., Marma, M.S., Lee, C.-Y., Kamal-Bahl, S., Kim, E.S., McKenna, C.E., 2010. Applied Physics Letters 96, 1237021–1237025.