NANO LETTERS
TiO2 Nanowire Bundle Microelectrode Based Impedance Immunosensor for Rapid and Sensitive Detection of Listeria monocytogenes
2008 Vol. 8, No. 9 2625-2631
Ronghui Wang,† Chuanmin Ruan,† Damira Kanayeva,‡ Kentu Lassiter,§ and Yanbin Li*,†,‡,§ Department of Biological and Agricultural Engineering, Cell and Molecular Biology Program, Center of Excellence for Poultry Science, UniVersity of Arkansas, FayetteVille, Arkansas 72701 Received February 5, 2008; Revised Manuscript Received June 2, 2008
ABSTRACT A novel TiO2 nanowire bundle microelectrode based immunosensor was demonstrated as a more sensitive, specific, and rapid technology for detection of Listeria monocytogenes. TiO2 nanowire bundle was prepared through a hydrothermal reaction of alkali with TiO2 powder and connected to gold microelectrodes with mask welding. Monoclonal antibodies were immobilized on the surface of a TiO2 nanowire bundle to specifically capture L. monocytogenes. Impedance change caused by the nanowire-antibody-bacteria complex was measured and correlated to bacterial number. This nanowire bundle based immunosensor could detect as low as 102 cfu/ml of L. monocytogenes in 1 h without significant interference from other foodborne pathogens.
Foodborne bacterial pathogens are a serious health threat worldwide. Major foodborne pathogens include Salmonella Typhimurium, Listeria monocytogenes, Staphylococcus aureus, Campylobacter jejuni, Clostridium perfringens, Yersinia enterocolitica, and Escherichia coli O157:H7.1 It has been reported that each year at least 76 million people in the United States become sick from foodborne pathogens,2 and such illnesses result in 325 000 hospitalizations and 5000 deaths annually.3 According to estimates by the USDA (United States Department of Agriculture),4 the medical costs and productivity losses resulting from foodborne illnesses stemming from these major food pathogens range between $6.5 billion and $34.9 billion annually. Most conventional methods such as cell culture, microscopy, biochemical tests, and luminescence are labor-intensive and time-consuming, often requiring 24∼48 h to obtain results. ELISA (enzyme-linked immunosorbent assay) and PCR (polymerase chain reaction), the two most popular technologies, may greatly reduce the assay time compared with traditional culture techniques. However, they still take several hours to identify target pathogens and lack the ability to detect bacteria in real time. Biosensors based on a * Corresponding author. Phone: 479-575-2424. Fax: 479-575-7139. E-mail:
[email protected]. † Department of Biological and Agricultural Engineering. ‡ Cell and Molecular Biology Program. § Center of Excellence for Poultry Science. 10.1021/nl080366q CCC: $40.75 Published on Web 08/21/2008
2008 American Chemical Society
combination of biological receptors and physical or chemical transducers present a unique technology with great potential to meet the need for the rapid, online, real time detection of biological agents.5,6 The immunological based biosensors or immunosensors (based on antibody-antigen affinity reaction) are among the most promising because of their high specificity and versatility in the detection of bacteria. Examples are amperometric immunosensors,7,8 potentiometric immunosensors,9 electrical impedance biosensors,10-12 piezoelectric biosensors,13,14 and optical biosensors.15-17 The reported immunosensors typically have a detection limit of 103∼106 cfu/ml and a detection time of 0.2∼2 h. Obviously, sensitivity is the major limitation and needs to be improved for most immunosensors. Nanowires/nanotubes have attracted much attention recently, especially in biosensing technologies. This is due to their unique semiconductive properties associated with the nanostructures, and they are believed to be ultrasensitive in performing single molecule sensing.18 TiO2 based semiconductive nanowires/nanotubes can be highly oriented on substrates or form free-standing membranes, either of which can be easily mounted on electrodes under an optical microscope.19,20 These TiO2 nanowires/nanotubes can have their band gaps systematically varied between 1.8 and 4.1 eV, which are not available for many other nanowires/ nanotubes. Such semiconductive one-dimensional nanostruc-
Figure 1. Long TiO2 nanowire bundle.
tures would be very useful for giving the best sensitivity ranges. In addition, the nanostructured TiO2 is usually easy to fabricate, has a large specific surface area, good biocompatibility, good chemical and photochemical stabilities, and negligible protein denaturation.21-24 These properties make nanostructured TiO2 an ideal sensing material. In this study, TiO2 nanowire bundle microelectrode based impedance immunosensor was demonstrated for more rapid and sensitive detection of Listeria monocytogenes, which is the cause of many recent well-publicized food poisoning outbreaks25,26 and responsible for serious infections in immunocompromised individuals and pregnant women. To the best of our knowledge, this is the first example to employ a TiO2 nanowire bundle in an impedance immunosensor for bacteria detection. The nanowire bundle microelectrodes were provided by Dr. Wenjun Dong and Dr. Ryan Tian at the University of Arkansas, Fayetteville, AR. The fabrication of TiO2 nanowire bundle microelectrode included two steps as described in the study by Tian et al.27 The first step was the synthesis of the TiO2 nanowire bundle. In a typical synthesis, 0.20 g of TiO2 powder (Degussa P25) was introduced into 40 mL of 10 mol/L alkali solution in a 150 mL Teflon-lined autoclave container. A cylindrical Teflon rod (3 cm in diameter and 4 cm in height) was set in the liner as the template. After 20 min sonication, the hydrothermal reaction was placed in an oven for 10 days at temperatures above 180 °C. The white, paper-like flexible nanowire bundle product was collected, washed with water, dried in air, and then separated with tweezers. The nanowire bundles have a diameter of 1∼20 µm (approximately 200∼2000 nanowires in each bundle), and the single nanowire’s diameter is between 60∼80 nm. The length of the nanowire bundle varies from 1 mm to several centimeters (Figure 1). The second step was the connection of the TiO2 nanowire bundle with a gold microelectrode. Under optical microscopy, the TiO2 nanowire bundle was placed on a glass substrate and then a 65 µm parallel Cu mask was laid over the nanowire bundle. Au sputtering (SC7260 Au sputter) coating (5 min) was then used to connect the crossed nanowire bundle with gold microelectrodes. Figure 2 shows the gold microelectrode crossed with the TiO2 nanowire bundle (a) and the gold microelectrode without TiO2 nanowire bundle used as a control (b). The whole analytical procedure contains three steps: gold electrode surface blocking, antibody immobilization, and bacteria detection. For gold surface blocking, the microelectrode was first covered with 99.5% ethanol for 3 min, rinsed with milliQ water, then incubated with 99% 2-methyl-2propanethiol for 20 min at room temperature. Following the incubation, the electrode was rinsed with 99.5% ethanol, milliQ water, then incubated with milliQ water for an 2626
Figure 2. (a) Gold microelectrode crossed with TiO2 nanowire bundle. (b) Gold microelectrode without TiO2 nanowire bundle used as a control.
additional 20 min at room temperature. For material safety consideration, the blocking procedure was performed in a chemical hood and the 2-methyl-2-propanethiol waste was collected in a marked waste container. Monoclonal antibodies specific for L. monocytogenes (420 µg/ml in PBS pH 7.4) were immobilized by directly incubating a drop (15 µL) of antibody on the TiO2 nanowire bundle surface at room temperature for 2 h, then rinsed with milliQ water to remove unbound antibody and dried with a stream of nitrogen gas. After the antibody was immobilized onto the TiO2 nanowire bundle surface, a 15 µL drop of bacteria was incubated on the surface of the TiO2 nanowire bundle for 50 min. Impedance measurements were performed in the presence of 0.1 M mannitol (12 µL) using an IM-6 impedance analyzer (BAS, West Lafayette, IN) with IM-6/THALES software. For all impedance measurements, a sine-modulated ac potential of 5 mV was applied across the TiO2 nanowire bundle microelectrodes and the magnitude of impedance and phase angle were measured in the frequency range of 1 Hz to 1 MHz. Further experimental details (reagents, bacteria, and cultures used in the tests) and the Dot Blot analysis procedure are given in Supporting Information. Immuno-Dot Blot analysis was used as a conventional immunoassay method for comparison with the biosensor method, and to observe the quality of the anti-L. monocytogenes antibodies used in the immunosensor. Seven concentrations of L. monocytogenes (2.2 × 102 to 2.2 × 108 cfu/ml) were assayed with the Dot Blot system, and the results are shown in Figure 3. No detectable signal was observed with concentrations of 102∼104 cfu/ml, and only very slight dots were seen at 105 cfu/ml. As expected, increasing the bacteria concentration (105∼108 cfu/ml) resulted in an increase of dot size and intensity. The results indicate that the purchased anti-L. monocytogenes antibody allows the detection of L. monocytogenes. The detection limit of this immunoassay Dot Blot analysis is 2.2 × 105 cfu/ml of pure cultures of L. monocytogenes. Figure 4 illustrates the principle for detection of L. monocytogenes using the TiO2 nanowire bundle microelectrode based impedance immunosensor. In order to avoid gold surface adsorption during antibody immobilization, the gold electrodes were blocked with 2-methyl-2-propanethiol (SH-(CH2)3-CH3), and only the TiO2 nanowire bundle surface was available for antibody binding. Since the semiconductive titanium dioxide (TiO2) nanowire carries Nano Lett., Vol. 8, No. 9, 2008
Figure 3. Detection of Listeria monocytogenes based on immunoassay Dot Blot analysis.
negative charges on their surface at pH > 6,28 the antibody immobilization was carried out in PBS buffer with pH at 7.4. The antibodies have a basic isoelectric point, and pH 7.4 was much below the isoelectric point of the antibodies, hence, the antibodies have positive charges and will directly bind to the TiO2 nanowire bundle surface. In this study, antiL. monocytogenes monoclonal antibodies were modified on the TiO2 nanowire bundle surface to specifically capture target bacteria. Then, changes in the impedance of the nanowire bundle-antibody-bacteria cell complex were measured and correlated to the presence of target L. monocytogenes in the sample. Figure 5 shows some typical impedance spectra of the TiO2 nanowire bundle microelectrode based impedance immunosensor for antibody immobilization and bacteria detection. Two TiO2 nanowire bundle based microelectrodes were subjected to the same treatment (gold surface blocking and antibody immobilization). One microelectrode was used to detect only BHI growth medium without bacteria (control), and the other was applied for detection of target L. monocytogenes. As can be seen, a significant decrease in impedance was observed after antibody immobilization for both microelectrodes, which indicated the successful modification of the TiO2 nanowire bundle surface with the antibody. For sample detection, a clear increase in impedance was seen for target L. monocytogenes (4.65 × 103 cfu/ml; Figure 5b), which has a maximum impedance change (83.6 kΩ) at frequency 1.04 kHz, while the control sample (growth medium without bacteria) only gave a negligible impedance change (Figure 5a). The change in impedance for antibody immobilization and L. monocytogenes detection were due to the binding of charged macromolecules or cells. Cui et al.29 and Hahm and Leigher30 used a silicon nanowires based sensor for direct electrical detection of biomolecule binding and showed that the binding of charged macromolecules could cause significant conductance change. In our experiments, impedance measurements were performed in the presence of 0.1 M mannitol, which has pH ∼ 7; hence, antibodies were highly positively charged, but the L. monocytogenes cell surface was highly negatively charged.31 Obviously, opposite impedNano Lett., Vol. 8, No. 9, 2008
ance behavior could be observed in our study; that is, the impedance decreased for antibody immobilization but increased for L. monocytogenes cells binding. Changes in impedance for TiO2 nanowire bundle microelectrode before and after biological binding can be further explained using the fitting values of the equivalent circuit elements listed in Table 1. It was found that the impedance data of the microelectrode chip, anti-Listeria antibody modified chip, and Listeria bound chip can be well-fitted using Randles’ equivalent circuit (Figure 6). The error of the impedance fittings is 1.9%, 2.7%, and 2.3%, respectively, for the three different chips using the software with the impedance analyzer. The equivalent circuit consists of the Ohmic resistance of the electrolyte (Rs), the polarization resistance (Rp), Warburg impedance (W), and the double layer capacitance (Cdl). Ohmic resistances of the electrolyte (Rs) for the microelectrode chip, anti-Listeria antibody modified chip, and Listeria bound chip are very small (
