4 CHAPTER Membrane-based electrochemical nanobiosensor for detection of virus 4.1 INTRODUCTION 4.1.1 West Nile Virus Domain III West Nile Virus (WNV) was first identified from a febrile female adult in the West Nile region of Uganda in 1937 [1]. WNV, a single-stranded positive sense RNA envelope virus, belongs to the Flaviviridae virus family. WNV has wide range of animal hosts and is transmitted mainly by a mosquito vector. WNV causes human illness that can progress to paralysis, encephalitis and death. WNV contains structural proteins. They are large envelope (E) protein, a single nucleocapsid protein (C) and a lipid membrane protein (M). The E protein monomer folds into domains: domain I, II and III. Domain I has an 93 eight-stranded β-barrel which takes part in the conformational changes associated with the acidification in the endosome. Domain II is 12β-strands and plays roles in dimerization, trimerization and fusion. Domain III (DIII) adopts an immunoglobulin-like fold, and contains surface exposed loops in the A B 4.8 nm Fig. 4.1 (A) Homology model of a WNV domain III protein. (B) Transmission electron micrograph of WNV particles obtained using Philips 208s transmission electron microscope. Sample was stained with phosphotungstic acid for min, rinsed with water, and dried before viewing. mature virion [2]. The sequence of WNV-DIII was well-studied [3]; it has short length of ca. 100 amino acids [4]. Some previous studies suggest that domain III could mediate flavivirus attachment to the host cell [5]. The envelope domain III protein is chosen in this work because of its important role in the implications in virulence. Immunoglobulin M antibody raised against domain III protein is used as the specific biorecognition probe for both 94 the WNV-DIII protein and WNV particle. Fig. 4.1 shows the structure of WNV-DIII protein and WNV particle. 4.1.2 Current analysis method for West Nile Virus detection In general, two types of tests have been developed for the diagnosis and screening of WNV. They are based on either antibodies (immunoglobulin M (IgM) or immunoglobulin G (IgG)) or nucleic acid tests in body fluid or tissue. Some methods for detections of WNV antibodies are complement fixation test (CF), hemagglutination- inhibition test (HI), plaque-reduction neutralization (PRNT), immune-fluorescence assay (IFA), enzyme-linked immunosorbent assay (ELISA), and the microsphere immunoassay (MIA). ELISA and MIA give high sensitivity and selectivity for WNV detection test, but it takes two or three days for the whole assay procedure while the IFA assays requires additional serum test and faces the cross-reactivity of WNV antibodies [6]. The other test for WNV diagnosis is the detection of WNV nucleic acid. Because virus presents in blood or plasma samples of infected patients at very low concentration, an in vitro amplification procedure is needed to increase WNV genetic material. Several amplification tests were found to enhance the 95 WNV detection such as the real-time polymerase chain reaction (PCR) test, which takes few hours and has extremely low detection limit for WNV in blood and plasma samples such as 0.1 plaque-forming units (pfu) of virus [7]. 4.1.3 Immunoglobulin M antibody During acute infection, serological WNV IgM antibody detection was found to be of higher diagnostic sensitivity than viral RNA determination by reverse transcriptase PCR. Moreover, immunoglobulin M (IgM) antibodies are less cross-reactive, whereas immunoglobulin G (IgG) antibodies show more cross-reactions with other flaviviruses [8]. Since in our biosensor, the sensing signal is based on the specific antigen-antibody binding, the IgM was chosen as the biorecognition element and was immobilized within the nanochannels. Immunoglobulin (Ig) is a globutin-type protein found in serum or other body fluids which possesses antibody activity. Each Ig unit is built up from light (L) and one heavy (H) polypeptide chains linked together by disulfide bonds (Fig. 4.2(A)). Based on antigenic and structural differences in the H chains, Immunoglobulin are divided into five classes of A, G, D, M and E. In the five immunoglobulin classes, immunoglobulin M is the largest antibody 96 and is the first antibody that appears when the body is challenged by the antigen. IgM normally exists as pentamer (Fig. 4.2(B)) but it can also exist as a monomer. In the pentamer form, IgM has a molecular mass of approximately 900 kD. A B Antigen binding sites Variable region on heavy chain Light chain S S S S Heavy chain Variable region on light chain Constant region on light chain Join chain Constant region on heavy chain Fig. 4.2 (A) Structural region of antibody molecule. (B) Structure of IgM in the pentamer form. In this work, the monoclonal IgM antibody (H5.46) which specifically binds to E protein of WNV was employed [9] to form the interaction with the WNV proteins and viral particles during the sensing process. 4.1.4 Immunosensor Immunosensor is the most specific biosensors in which the biorecognition elements are antibodies (Ab). Immunosensors provide low detection limits and can be applied in wide range of substances. In immunosensors, antigen-antibody interactions are transduced directly into 97 physical signals to sense antigen (Ag). Ab + Ag ' Ab-Ag In a typical immunosensor, antibodies which are the globular protein produced by organisms to bind foreign molecules (antigens) and mark them for elimination from the organism, are linked on a stable solid support and coupled to a transducer element. Prepared antibody can be monoclonal or polyclonal; the latter is cheaper but possesses varied binding affinities due to many different types of biorecognition elements present, with poorly reproducible proportions across different preparations. The design and preparation of an optimum interface between the biorecognition elements and the transducer material are the key part of biosensor development. Electrochemical immunosensors that combine specific immunoreactions with an appropriate electrochemical transduction have increased interests due to low cost, rapid response and simple-to-use procedure. There are two categories of electrochemical immunosensors, one detects antigen directly, while the other senses indirectly. In the indirect electrochemical immunosensors, the signal transduction is produced by the 98 second compound or reaction (such as mediators). Our immunosensor in this study is based on the indirect sensing mechanism. 4.1.5 Nanoporous membrane based biosensors Nanoporous membranes, comprising uniform and regularly spaced channels of nanometer dimension, have attracted great interest as a template material for the incorporation of various materials including biological molecules, metals, semiconductors, and polymers within the nanosized channels [10-12]. Nanoporous membranes possess uniform pore sizes, high aspect ratio, and high surface areas, are relatively easy to prepare, and are inexpensive by comparison to conventional lithographic techniques. In general, incorporation of interesting and useful materials within the membrane channels impart new physicochemical or biological properties which results in new or improved membrane applications. In particular, electrochemical nanobiosensors using nanoporous membranes, such as a nanoporous semiconductor [13], porous conducting polymer [14], track-etched polymer [15, 16], and porous alumina[17, 18], have been reported. Electrochemical biosensors generally involve redox enzymes for the 99 conversion of substrate into product [19]. The response signal at the sensing electrode is derived from the redox enzyme directly by electrically conducting linkers [20] or indirectly via redox mediator or redox reaction of the substrate or product [21]. More recently, non redox antibodies and single-stranded nucleic acids are promising biorecognition molecules because of high recognition specificity for the analyte of interest and are applicable to large numbers of nonredox active biological analytes, as reported by Wang et al. and other researchers [22-26]. Electrochemical sensing can be derived from the changes in physical proximity of the redox labeled biorecognition molecules to the sensing electrode [24, 25] or variation in the concentration of the electrochemical tags close to the sensing electrode, before and after binding to the analytes of interest [26] or via the gold nanoparticle amplification method [22, 23]. Until now, several methods using multiarray nanopores (or nanochannels) embedded within membranes have demonstrated that rapid analysis is potentially suitable for extreme analysis of a small sample quantity. For example, using an array of conical gold nanopores functionalized with thiolated-biotin [16], specific detection of streptavidin modified analytes can be achieved. Silicon dioxide based nanopores are 100 modified using silane chemistry to influence transport of charged analytes via electrostatic interactions between the charged analytes and the charged functional groups at the modified surface [27]. A recent interesting report uses the change in protein structure upon binding to target analyte to influence the electrolyte conductivity within Au-coated nanopores [15]. Unlike the impedance or conductance measurement methods, we use a highly sensitive differential pulse voltammetry method to monitor the Faradaic current of redox species at the membrane-electrode interface which generates the biosensor signal in response to antigenic analytes bound within the multiarray membrane nanochannels. In addition, the sensing electrode is directly coated onto the nanoporous membrane, comprising multiarray nanochannels in order to couple the mass transport rate of the redox species within the membrane to the electrochemical reaction at the membrane-electrode interface. Previous work using a nanoporous alumina membrane [18] suggested that it is ideally suited for this work because of high pore density which offers high current flux, tunable pore sizes, rigid support structure, chemical and thermal stability, and ease of preparation using electrochemical anodization of aluminum. In this work, we report the nanoporous membrane-based biosensor that 101 can be used to detect extremely low concentration of a virus protein and particle using a rapid sensing time of 30 min. The nanobiosensor response is first optimized using the West Nile virus-domain III protein to which the immunoglobulin M (IgM) biorecognition probe binds to, with subsequent application for the direct detection of the West Nile virus (WNV) particle. AC voltammetry reveals that contribution of diffusion is significant in the observed reduced biosensor response signal toward ferrocenemethanol, in the presence of the virus protein or the particle. Equilibrium constants for the antigen-antibody binding are derived from the electrochemical response signal using simple Langmuir isotherm. Limits of detection for the viral protein and particles are pg mL-1 and viral particles per 100 mL, respectively, which are similar to detection limits of viruses using PCR techniques [28, 29]. Finally, we demonstrate the highly specific sensing of the WNV particle is readily achieved in a complex medium, blood serum containing other proteins. 4.2 EXPERIMENT 4.2.1 Reagents and Materials Bovine serum albumin (BSA, >98%), ferrocenemethanol (FeMeOH, >99%), sodium dihydrogenphosphate dehydrate, phosphoric acid and 102 Normalized current (I WNV-DIII/I WNV-DIII=0) 1.5 A 30 60 1.0 0.5 0.0 10 20 30 40 50 60 -1 Concentration of WNV-DIII protein (pg mL ) B Normalized current (I WNV-DIII/I WNV-DIII=0) 1.5 30 60 1.0 0.5 0.0 0.0 0.2 0.4 0.6 -1 Concentration of WNV particle (PFU mL ) Fig. 4.8 Biosensor response signal towards (A) WNV-DIII protein and (B) viral particle in pH 6.8, 1.7 M NaCl. Biosensors were prepared from membrane electrodes etched for 0, 30, 60 in 3% phosphoric acid solution and using 0.2 μg L-1 IgM and 200 μg L-1 BSA. Lines are non-linear curve fits using Eqn. 4.2. To demonstrate the usefulness of this method for early detection of the WNV disease, it is necessary for the biosensor to detect whole WNV particle instead of WNV-DIII protein which not occur in its free form in infected 120 patients. Several techniques are capable of sensing nanosized particles [38], but specific identification and characterization of the nature of nanosized particles remain a challenge. Fig. 4.8A shows the biosensor response signal towards WNV particle for different membrane electrodes etched over different time durations. Interestingly, the unetched (0 min) alumina membrane electrode responds to the large WNV particle (ca. 50 nm) which are significantly larger than the average 10-20 nm diameter pore openings of the membrane (Fig. 4.8B). Unlike WNV-DIII protein, the biosensor response towards WNV particle is stable, likely because one virion binds to several IgM molecules positioned close to or at the membrane pore openings. It is known that the surface of one virion contains 180 copies of membrane envelope glycoprotein from which WNV-DIII is derived [39]. One particularly useful characteristic of the biosensor is that highest biosensor sensitivities towards both WNV-DIII protein and WNV particle are achieved using membrane electrodes etched over long time. Thus, we are able to prepare biosensors using identical optimal conditions, to give highest sensitivity response towards both WNV-DIII protein and WNV particle, though the size ratio of the analytes is 1:10. This can only be explained by the nanochannels having 121 varied diameters along the lengths, with dimensions appropriate for the entrapment of both large viral particle and small protein antigen. We have shown in previous work that the alumina membrane electrode when etched chemically, produced wider nanochannels at the exterior part of membrane (>70 nm), while maintaining a narrow size of ca. 30 nm for most channels close to the electrode surface [18]. If the sensing of WNV-DIII and WNV particle rely on specific antigen-antibody binding and our experiment conditions allow near equilibrium binding before measuring the response signal, then the overall biosensor response signal can be expressed as a function of the antigen-antibody binding equilibrium constant K. In the following, we describe the biosensor response using simple Langmuir isotherm to derive K values for WNV-DIII-IgM and WNV particle-IgM immunocomplexes. 4.3.4 Antibody-antigen binding Basically, antigen-antibody interactions are the bimolecular reaction that can be described using the following simple equilibrium equation: Ag + Ab ' Ag-Ab Eqn. 4.1 where K= kf/kb is the binding equilibrium constant for the antigen (Ag) and antibody (Ab) interaction. 122 If antigen-antibody binding occurs within an IgM-adsorbed nanochannel, the normalized biosensor response signal (IWNV/IWNV=0) will decrease from because of reduced mass transport rate of ferrocenemethanol moving through the blocked nanochannels to the underlying sensing electrode. The binding of antigen to IgM within each IgM-adsorbed nanochannel is expected to occur independently of other antigen-IgM binding in the same nanochannel or other nanochannels. Each monomeric unit within the IgM pentamer is regarded as one binding site and we assume at equilibrium, every antigen-IgM monomer complex formed in the nanochannels is equivalent when measured by the DPV method. Thus, surface coverage of antigen molecules at the IgM binding sites (θ) for a particular antigen concentration relates directly to the normalized biosensor response signal (IWNV/IWNV=0) as θ = {1 − ( I WNV / IWNV =0 )} . Surface coverage of antigen at these effective binding sites (θ) can be correlated to antigen concentration ([A]) and equilibrium constant for the specific binding between antibody and antigen (K) according to Langmuir isotherm, as follows: - I WNV /I WNV =0 θ max = K [A] + K [A] Eqn. 4.2 where θ and θmax preclude binding sites at external surface of membrane which bind to antigen but not influence the overall biosensor response 123 signal; K is antigen-antibody binding equilibrium constant. {1- (IWNV/IWNV=0)} (or θ) is normalized against the maximum surface coverage θmax at high antigen concentrations to provide for differences in IgM loading or structure of the membrane nanochannels under different conditions. Using Eqn. 4.2, very good fits between the curve fitted and experimental data are obtained for WNV-DIII and WNV particle, as shown in Fig. 4.7, 4.7, 4.8. Under optimal conditions, the best fitted values for different biosensors give K = (6.4 ± 1)×1011 M-1 and θmax = 1.1 ± 0.1 for WNV-DIII and K = (4.3 ± 0.7)×1016 M-1 and θmax = 0.6 ± 0.0 for WNV particle (Table 4.1). θmax for WNV-DIII is ca. but less than for WNV particles because of poorer fit of the larger WNV particles within the IgM adsorbed nanochannels. K values for WNV-DIII protein-IgM binding fall within expected range of antigen-antibody binding affinity from 106-1012 M-1[40]. At 3.5 M NaCl, the K value for WNV-DIII protein decreases to low range of 8.47×109 M-1, attributed to the influence of high ionic strength on the antigen-antibody binding affinity previously discussed in Fig 4.6(F). Interestingly for WNV particle, consistent K values of fold magnitude larger than those for protein antigen are obtained. These large binding affinity K values are attributed to multivalent binding effect [40] between one virus particle containing multiple copies of the WNV-DIII protein and several IgM molecules. 124 Table 4.1 Best fitted data for binding affinity K and maximum surface coverage θmax between IgM and WNV-DIII protein or WNV particle, calculated from Eqn. 4.2. WNV-DIII Protein Parameter investigated IgM concentration pH of buffered electrolyte Ionic strength of buffered electrolyte Etching time Reproducibility Variations K (1011 M-1) θmax 0.1 µM 2.24 ± 1.91 1.00 ± 0.41 0.2 µM 4.89 ± 0.62 1.16 ± 0.06 0.4 µM 5.83 ± 1.13 1.32 ± 0.11 pH 8.2 6.48 ± 1.18 0.82 ± 0.05 pH 7.6 2.77 ± 0.52 1.57 ± 0.16 pH 6.8 4.9 ± 0.6 1.16 ± 0.06 pH 6.2 2.48 ± 0.32 1.29 ± 0.09 0.1M 2.48 ± 0.32 1.29 ± 0.09 1.7M 4.89 ± 0.63 1.16 ± 0.06 3.5M 0.08 ± 0.25 3.99 ± 11.4 1.75 ± 3.14 1.00 ± 1.12 30 3.89 ± 0.30 1.04 ± 0.04 60 4.18 ± 0.43 1.26 ± 0.06 biosensors 6.40 ± 1.40 1.12 ± 0.1 WNV Particle Parameter investigated Variations K (1016 M-1) θmax 3.67 ± 7.45 0.34 ± 0.02 30 4.05 ± 0.92 0.53 ± 0.04 60 3.50 ± 0.70 0.72 ± 0.05 Reproducibility biosensors 4.25 ± 0.73 0.62 ± 0.03 Real Sample Test In phosphate buffer 5.92 ± 0.71 0.61 ± 0.02 In serum 3.50 ± 0.70 0.73 ± 0.04 Etching time 125 4.3.5 Analytical performance of the biosensor. The detection of the viral protein and particle are logarithmically linear up to 53 pg mL-1 (R2 = 0.99) and 50 viral particles per 100 mL-1 (R2 = 0.93) in pH 7, with extremely low detection limits of pg mL-1 and ca. PFU mL-1. The linear logarithmic concentration dependence range is from 4-53 pg mL-1 (R2 = 0.99) for WNV-DIII and from 0.02-0.5 PFU mL-1 (R2 = 0.93) for WNV particles. In addition, the membrane-based nanobiosensor requires only ca. µl of 0.2 µg L-1 antibody during the biosensor preparation procedure, unlike current immunoassay methods which require large quantity of antibody (50 µL of 20 mg L-1). Overall, the responses of different biosensors towards WNV-DIII or WNV particle give R.S.D. of 1-7% or 2-9% respectively. 4.3.6 Detection of WNV particle in Blood Serum. Because sensor techniques based on immunoassays detecting immunoresponse is effective a few days after viral infection, early detection of disease before onset of symptoms is difficult [41]. An alternative approach is to detect the virus directly without relying on antibody detection. To evaluate the potential application of our biosensor for rapid and early detection of viral particles in diseased patients using minimal sample preparation, it is necessary to demonstrate measurements in blood serum samples. 126 Normalized current (I WNV-DIII/I WNV-DIII=0) 1.0 Normalized current (I WNV-DIII/I WNV-DIII=0) 1.2 1.0 0.8 0.8 0.6 0.4 0.2 0.0 -2.0 0.6 -1.5 -1.0 -0.5 0.0 Log( [WNV particle]) 0.4 In blood serum In phosphate buffer 0.2 0.0 0.0 0.2 0.4 0.6 -1 Concentration of WNV particle (PFU mL ) Fig. 4.9 Comparison of biosensor response signal towards WNV particle in 0.1 M phosphate buffer (pH 6.8, ionic strength of 1.7 M) and in blood serum, using same sensor preparation procedure (see experimental section). Inset: Plot of normalized current vs. log ([WNV particles]) for the sensing of inactivated WNV particle in blood serum. Fig. 4.9 shows the change in response signals of the biosensor in blood serum and phosphate buffer samples spiked with different amounts of WNV particle. Paired t-test performed on the data derived in serum and buffer samples reveal excellent correlation in which the t statistic (0.49) is lower than the critical value of t even at the 50% confidence level. Standard errors of biosensor response signals obtained in serum samples are between 5-11%, higher than that obtained in buffer samples (1-6%). To illustrate the usefulness of this 127 method, a blood serum sample was prepared with 0.030 PFU mL-1 inactivated WNV particles. Using standard addition method, the blood sample was spiked with concentrations of WNV particle and the measured value by the biosensor was 0.029 + 0.002 PFU mL-1. The t-test was employed to compare the measured mean against 0.030 + 0.001 PFU mL-1 for the preparation procedure with errors arising from the micro-pipette. The t value obtained was below the critical value of t at the 50 % confidence level which indicates signal interference from other serum proteins can be considered insignificant. 128 4.4 CONCLUSIONS We report a new membrane-based electrochemical nanobiosensor for the sensing of a virus envelope protein (West Nile Virus Domain III protein), made from a nanoporous alumina membrane laid over a sensing electrode. A low detection limit of pg mL-1 is easily achieved using voltammetry measurement method. In addition, the biosensor shows exceptionally high sensitivity towards whole WNV particle with an extremely low detection limit of 0.02 PFU mL-1 which is comparable to polymerase chain reaction techniques. AC voltammetry studies reveal the biosensor response signal depends on the mass transport of the redox ferrocenemethanol moving through the nanochannels of the membrane towards the underlying sensing electrode. The superior biosensor performance of quantitating whole viral particles specifically without additional sample treatment steps is demonstrated using spiked blood serum. We have succeeded in using the Langmuir-isothermal equation to calculate the antibody-antigen binding constant. Loading of other immunoglobulin antibodies in the nanoporous membrane layer can further extend the biosensor use for unlimited sensing of other proteins and viruses with excellent performance. 129 REFERENCES 1. Mukhopadhyay, S., B.S. Kim, P.R. Chipman, M.G. Rossmann, and R.J. Kuhn, Science, 2003. 302(5643): p. 248-248. 2. Oliphant, T., M. Engle, G.E. Nybakken, C. Doane, S. Johnson, L. Huang, S. Gorlatov, E. Mehlhop, A. Marri, K.M. Chung, G.D. Ebel, L.D. Kramer, D.H. Fremont, and M.S. Diamond, Nature Medicine, 2005. 11(5): p. 522-530. 3. Kanai, R., K. Kar, K. Anthony, L.H. Gould, M. Ledizet, E. Fikrig, W.A. Marasco, R.A. Koski, and Y. Modis, Journal of Virology, 2006. 80(22): p. 11000-11008. 4. Volk, D.E., D.W.C. Beasley, D.A. Kallick, M.R. Holbrook, A.D.T. Barrett, and D.G. Gorenstein, Journal of Biological Chemistry, 2004. 279(37): p. 38755-38761. 5. Anderson, R., Manipulation of cell surface macromolecules by flaviviruses, in Advances in Virus Research. 2003. p. 229-274. 6. Ionescu, R.E., S. Cosnier, S. Herrmann, and R.S. Marks, Analytical Chemistry, 2007. 79(22): p. 8662-8668. 7. 130 Lanciotti, R.S. and A.J. Kerst, Journal of Clinical Microbiology, 2001. 39(12): p. 4506-4513. 8. Tardei, G., S. Ruta, V. Chitu, C. Rossi, T.F. Tsai, and C. Cernescu, Journal of Clinical Microbiology, 2000. 38(6): p. 2232-2239. 9. Lanciotti, R.S., J.T. Roehrig, V. Deubel, J. Smith, M. Parker, K. Steele, B. Crise, K.E. Volpe, M.B. Crabtree, J.H. Scherret, R.A. Hall, J.S. MacKenzie, C.B. Cropp, B. Panigrahy, E. Ostlund, B. Schmitt, M. Malkinson, C. Banet, J. Weissman, N. Komar, H.M. Savage, W. Stone, T. McNamara, and D.J. Gubler, Science, 1999. 286(5448): p. 2333-2337. 10. Lakshmi, B.B., P.K. Dorhout, and C.R. Martin, Chemistry of Materials, 1997. 9(3): p. 857-862. 11. Masuda, H. and K. Fukuda, Science, 1995. 268(5216): p. 1466-1468. 12. Huczko, A., Applied Physics A: Materials Science and Processing, 2000. 70(4): p. 365-376. 13. Yang, M.L., J. Wang, H.Q. Li, J.G. Zheng, and N.Q.N. Wu, Nanotechnology, 2008. 19(7): p. 6. 14. Ekanayake, E., D.M.G. Preethichandra, and K. Kaneto, Biosensors & Bioelectronics, 2007. 23(1): p. 107-113. 131 15. Tripathi, A., J. Wang, L.A. Luck, and Suni, II, Analytical Chemistry, 2007. 79(3): p. 1266-1270. 16. Sexton, L.T., L.P. Horne, and C.R. Martin, Molecular Biosystems, 2007. 3: p. 667-685. 17. Heilmann, A., N. Teuscher, A. Kiesow, D. Janasek, and U. Spohn, Journal of Nanoscience and Nanotechnology, 2003. 3(5): p. 375-379. 18. Koh, G., S. Agarwal, P.-S. Cheow, and C.-S. Toh, Electrochimica Acta, 2007. 53(2): p. 803-810. 19. Wang, J., Chemical Reviews, 2008. 108(2): p. 814-825. 20. Bartlett, P.N. and P.R. Birkin. The application of conducting polymers in biosensors. in Symposium H on Molecular Electronics: Doping and Recognition in Nanostructured Materials of the E-MRS Spring Conference. 1993. Strasbourg, France. 21. Rajesh, V. Bisht, W. Takashima, and K. Kaneto, Biomaterials, 2005. 26(17): p. 3683-3690. 22. Lin, Y.Y., J. Wang, G.D. Liu, H. Wu, C.M. Wai, and Y.H. Lin, Biosensors & Bioelectronics, 2008. 23(11): p. 1659-1665. 23. 132 Wang, J., W.Y. Meng, X.F. Zheng, S.L. Liu, and G.X. Li, Biosensors & Bioelectronics, 2009. 24(6): p. 1598-1602. 24. Radi, A.E., J.L.A. Sanchez, E. Baldrich, and C.K. O'Sullivan, Journal of the American Chemical Society, 2006. 128(1): p. 117-124. 25. Tan, E.S.Q., R. Wivanius, and C.-S. Toh, Electroanalysis, 2009. 21(6): p. 749-754. 26. Cai, H., T.M.H. Lee, and I.M. Hsing, Sensors and Actuators B-Chemical, 2006. 114(1): p. 433-437. 27. Kim, Y.-R., J. Min, I.-H. Lee, S. Kim, A.-G. Kim, K. Kim, K. Namkoong, and C. Ko, Biosensors and Bioelectronics, 2007. 22: p. 2926-2931. 28. Cordeya, S., Y. Thomas, P. Cherpillod, S. van Belle, C. Tapparel, and L. Kaiser, Journal of Virological Methods, 2009. 156(1-2): p. 166-168. 29. McKinney, M.D., S.J. Moon, D.A. Kulesh, T. Larsen, and R.J. Schoepp, Journal of Clinical Virology, 2009. 44(1): p. 39-42. 30. Chin, J.F.L., J.J.H. Chu, and M.L. Ng, Microbes and Infection, 2007. 9(1): p. 1-6. 31. Roberts, C.J., P.M. Williams, J. Davies, A.C. Dawkes, J. Sefton, J.C. Edwards, A.G. Haymes, C. Bestwick, M.C. Davies, and S.J.B. Tendler, 133 Langmuir, 1995. 11(5): p. 1822-1826. 32. Liang, H., C.F. Reich, D.S. Pisetsky, and P.E. Lipsky, Scandinavian Journal of Immunology, 2001. 54(6): p. 551-563. 33. Todorovi, N., S.K. Milonji, and V.T. Dondur, Materials Science Forum, 2004. 453-454: p. 361-366. 34. Wright, G.G. and V. Schomaker, Journal of the American Chemical Society, 1948. 70(1): p. 356-364. 35. Voet, D. and J.G. Voet, Biochemistry 1990, Wiley: New York. 36. Bruno, S., W. Gorczyca, and Z. Darzynkiewicz, Cytometry, 1992. 13(5): p. 496-501. 37. Sada, E., S. Katoh, K. Sukai, M. Tohma, and A. Kondo, Biotechnology and Bioengineering, 1986. 28(10): p. 1497-1502. 38. Park, K., J.S. Kim, and A.L. Miller, Journal of Nanoparticle Research, 2009. 11(1): p. 175-183. 39. Nybakken, G.E., T. Oliphant, S. Johnson, S. Burke, M.S. Diamond, and D.H. Fremont, Nature, 2005. 437(7059): p. 764-769. 40. Roitt, I., J. Brostoff, and D. Male, Immunology. 6th ed. 2001, New York: Mosby. 134 41. Hogrefe, W.R., R. Moore, M. Lape-Nixon, M. Wagner, and H.E. Prince, Journal of Clinical Microbiology, 2004. 42(10): p. 4641-4648. 135 [...]... response range Fig 4. 7C and Fig 4. 7D show the effect of pH on the response signal of the membrane electrode 4 different membrane electrodes were prepared in IgM solutions of different pHs and subsequent responses towards ferrocenemethanol were measured in the absence (Fig 4. 7C) and presence of antigens (Fig 4. 7D) By increasing 116 the immobilization pH from 6.2 to 7.6, the loading amount of IgM increases... sensitivity at 3.5 M NaCl 118 4. 3.3 Effect of membrane porous structure To optimize the membrane porous structure for the sensing of WNV-III, 3 different biosensors were prepared from membrane electrodes etched in phosphoric acid over time duration of 0, 30, 60 min before immobilization of IgM probe molecules We observe in Fig 4. 8A that the biosensor prepared from unetched (0 min) alumina membrane electrode gives... vs Ag/AgCl (V) Fig 4. 5 AC Voltammetry response of the membrane electrodes for the sensing of WNV-DIII protein 20 pg mL-1 at the frequency range from 0.7-5000 Hz, with 1.0 mM ferrocenemethanol in 0.1 M phosphate pH 6.8 buffer solution as the electrolyte 111 A 12.0 A (10 -4 cm2) 24. 9 ± 0.9 24. 5 ± 0.6 21.5 ± 0.6 17.3 ± 0 .4 14. 4 ± 0.3 10.0 I p (μA) 8.0 6.0 4. 0 Alumina 60min IgM WNV-DIII protein 7 pg/mL... 2.0 WNV particle 0.03 PFU/mL A (10 -4 cm2) I p (μA) WNV particle 0.05 PFU/mL 4. 61 ± 0.60 6.63 ± 0.50 4. 13 ± 0.18 1.60 ± 0.16 1.5 1.0 0.5 0.0 0.0 50.0 100.0 150.0 200.0 ω 1/2 (s-1/2) Fig 4. 6 Plot of AC voltammetry responses (Ip) of the membrane electrodes as a function of square root angular frequency (w1/2) after each step of the sensor preparation procedure for sensing of (A) WNV-DIII protein (B) WNV... provide for differences in IgM loading or structure of the membrane nanochannels under different conditions Using Eqn 4. 2, very good fits between the curve fitted and experimental data are obtained for WNV-DIII and WNV particle, as shown in Fig 4. 7, 4. 7, 4. 8 Under optimal conditions, the best fitted values for 6 different biosensors give K = (6 .4 ± 1)×1011 M-1 and θmax = 1.1 ± 0.1 for WNV-DIII and K = (4. 3... 0.6 ± 0.0 for WNV particle (Table 4. 1) θmax for WNV-DIII is ca 1 but less than 1 for WNV particles because of poorer fit of the larger WNV particles within the IgM adsorbed nanochannels K values for WNV-DIII protein-IgM binding fall within expected range of antigen-antibody binding affinity from 106-1012 M-1 [40 ] At 3.5 M NaCl, the K value for WNV-DIII protein decreases to low range of 8 .47 ×109 M-1,... 2 .48 ± 0.32 1.29 ± 0.09 0.1M 2 .48 ± 0.32 1.29 ± 0.09 1.7M 4. 89 ± 0.63 1.16 ± 0.06 3.5M 0.08 ± 0.25 3.99 ± 11 .4 1.75 ± 3. 14 1.00 ± 1.12 30 min 3.89 ± 0.30 1. 04 ± 0. 04 60 min Etching time 0.1 µM 0 min Ionic strength of buffered electrolyte θmax pH 6.2 pH of buffered electrolyte K (1011 M-1) pH 8.2 IgM concentration Variations 4. 18 ± 0 .43 1.26 ± 0.06 6 biosensors 6 .40 ± 1 .40 1.12 ± 0.1 WNV Particle Parameter... extrapolated to IWNV/IWNV=0 = 1 to derive the concentration of the unknown sample 4. 3 RESULT AND DISCUSSION 4. 3.1 West Nile Virus Domain III protein detection using membrane- based electrochemical biosensor Fig 4. 3 demonstrates the working principle of membrane- based electrochemical biosensor for WNV-DIII protein and particle accordingly the 107 specific binding of immunoglobulin M antibody with the WNV-DIII protein... antibody positioned along the walls of membrane nano-channels Fig 4. 4(A) shows the typical differential pulse voltammograms (DPV) of 108 ferrocenemethanol obtained at a membrane electrode immersed in a sensing solution containing redox active ferrocenemethanol, after each step of the biosensor preparation procedure Fig 4. 4B shows that the DPV response signal of the assembled membrane biosensor is sensitive... Prof Ng in collaboration 4. 2.2 Inactivation of virus Inactivation of virus was performed by the addition of binary ethyleneimine to the virus suspension to give a final concentration of 0.1 M The suspension was subsequently incubated at 37 °C for eight hours with constant shaking In order to validate for the complete inactivation of all infectious viral particles in the suspension, small aliquots of . pg/mL WNV-DIII protein 33 pg/mL U 24. 9 ± 0.9  24. 5 ± 0.6 c21.5 ± 0.6 17.3 ± 0 .4 ¼ 14. 4 ± 0.3 A (10 -4 cm 2 ) U 24. 9 ± 0.9  24. 5 ± 0.6 c21.5 ± 0.6 17.3 ± 0 .4 ¼ 14. 4 ± 0.3 A (10 -4 cm 2 ) B 0.0 0.5 1.0 1.5 2.0 2.5 0.0. 0.9  24. 5 ± 0.6 c21.5 ± 0.6 17.3 ± 0 .4 ¼ 14. 4 ± 0.3 A (10 -4 cm 2 ) U 24. 9 ± 0.9  24. 5 ± 0.6 c21.5 ± 0.6 17.3 ± 0 .4 ¼ 14. 4 ± 0.3 A (10 -4 cm 2 ) A 0.0 2.0 4. 0 6.0 8.0 10.0 12.0 0 50 100 150 200 ω 1/2 . PFU/mL U4.61 ± 0.60 6.63 ± 0.50 c4.13 ± 0.18 1.60 ± 0.16 A (10 -4 cm 2 ) U4.61 ± 0.60 6.63 ± 0.50 c4.13 ± 0.18 1.60 ± 0.16 A (10 -4 cm 2 ) Fig. 4. 6 Plot of AC voltammetry responses (I p ) of