A magnetoelastic bioaffinity-based sensor for avidin

A magnetoelastic bioaffinity-based sensor for avidin

Biosensors and Bioelectronics 19 (2004) 1695–1701 A magnetoelastic bioaffinity-based sensor for avidin Chuanmin Ruan, Kefeng Zeng, Oomman K. Varghese...

306KB Sizes 20 Downloads 51 Views

Biosensors and Bioelectronics 19 (2004) 1695–1701

A magnetoelastic bioaffinity-based sensor for avidin Chuanmin Ruan, Kefeng Zeng, Oomman K. Varghese, Craig A. Grimes∗ Departments of Electrical Engineering, Materials Science and Engineering, 217 Materials Research Laboratory, The Pennsylvania State University, University Park, Pennsylvania, PA 16802, USA Received 23 October 2003; received in revised form 17 December 2003; accepted 7 January 2004

Abstract A magnetoelastic bioaffinity sensor coupled with biocatalytic precipitation is described for avidin detection. The non-specific adsorption characteristics of streptavidin on different functionalized sensor surfaces are examined. It is found that a biotinylated poly(ethylene glycol) (PEG) interface can effectively block non-specific adsorption of proteins. Coupled with the PEG immobilized sensor surface, alkaline phosphatase (AP) labeled streptavidin is used to track specific binding on the sensor. This mass-change-based signal is amplified by the accumulation on the sensor of insoluble products of 5-bromo-4-chloro-3-indolyl phosphate catalyzed by AP. The resulting mass loading on the sensor surface in turn shifts the resonance frequency of the magnetoelastic sensors, with an avidin detection limit of approximately 200 ng/ml. © 2004 Elsevier B.V. All rights reserved. Keywords: Magnetoelastic sensor; Affinity sensor; Biotin–avidin; Biocatalytic reaction; Poly(ethylene glycol); Specific absorption; Specificity

1. Introduction Affinity-based biosensors are composed of a biological recognition element such as an antibody, receptor protein, biomimetric material, or DNA interfaced to a signal transducer, which together relate the concentration of an analyte to a measurable electronic signal (Rogers, 2000; Katz and Willner, 2003; Yoon et al., 2000). A variety of interface signal transduction technologies including optical (absorbance, luminescence, fluorescence, surface plasmon resonance), electrochemical (amperometric, impedimetric, potentiometric), and gravimetric (surface acoustic, quartz crystal microbalance) have been designed for affinity-based biosensors (Hanshoff et al., 2000; Whelan et al., 2002; Ruan et al., 2002). The signal transducer largely determines instrumentation cost, sensor cost, sensitivity boundaries for the target analyte, physical limitations such as size and portability, as well as data acquisition and signal processing characteristics. Magnetoelastic amorphous thick films are comprised of an alloy class that is largely iron, nickel, molybdenum and boron (Barandiaran and Gutierrez, 1997). A magnetic field impulse is used to impart energy into the ribbon-like mag∗ Corresponding author. Tel.: +1-814-865-9142; fax: +1-814-865-6780. E-mail address: [email protected] (C.A. Grimes).

0956-5663/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.bios.2004.01.005

netoelastic sensor, causing it to vibrate at a characteristic resonance frequency as illustrated schematically in Fig. 1. Fig. 2 shows both single sensors, as well as several different magnetoelastic sensor arrays. The starting magnetoelastic material is purchased in ribbon form, it is shipped wound on a reel, at cost of approximately US$ 500/kg from Honeywell Corporation. Starting with a 27 ␮m thick 2 mm wide ribbon from which sensors of 6 mm length are shear cut, one kg of material is sufficient for the fabrication of approximately 391,000 sensors. At US$ 500/kg, the material cost per sensor is US$ 0.0013. The cost of physically cutting a 2 mm wide sensor ranges from US$ 0.02 per sensor by laser cutting, to approximately US$ 0.001 for automated mechanical shear cutting. The monitoring electronics have a cost of approximately US$ 80. The low cost readily enables use of the sensors on a disposable basis. The resonance frequency of a magnetoelastic sensor shifts linearly in response to small mass loads, hence like quartz crystal microbalance (QCM) sensors the magnetoelastic sensor platform can be used for biological or chemical detection when coupled with appropriately sensitive coatings (Ersöz et al., 2002; Cai and Grimes, 2001). Since magnetoelastic sensors are excited by a magnetic field, and transduce information magnetically or acoustically, no direct wire connections to the sensor are necessary for signal telemetry. The response of a magnetoelastic sensor can also be used to characterize interfacial phenomena, such as the viscoelastic properties of an adhered

1696

C. Ruan et al. / Biosensors and Bioelectronics 19 (2004) 1695–1701 Pick-up coil, or microphone

Drive coil

Magnetoelastic sensor Field

Voltage

Time SEND

Amplitude

Time RECEIVE

Frequency CONVERT

Fig. 1. Drawing illustrating the remote query nature of the magnetoelastic sensor platform. The sensor is interrogated by a magnetic field impulse; the transient response is captured and then converted into the frequency domain using a fast Fourier transform. The resonant frequency is tracked to provide chemical and environmental information.

biomaterial or solution (Puckett et al., 2003; Schmidt and Grimes, 2001; Grimes et al., 2000), surface charges of adsorbed molecules (Chiriac et al., 1999), surface precipitation (Ruan et al., 2003) and surface roughness (Jain et al., 2001). Signal amplification of biological affinity events is a critical challenge to the future development of advanced affinity-based biosensors. Enzyme catalytic amplification is a commonly used means of signal amplification in biosensors and immunoassays. Ebersole et al. (1990) initially investigated an amplified mass immunosorbent assay, achieving a sandwich enzyme-linked immunosorbent assay on a microgravimetric quartz crystal microbalance device. Su and O’Shea (2001) recently expended this concept to develop hydrogen perxide and glucose sensors by quantifying the biocatalyzed precipitation using surface plasmon resonance. Katz and Willner (2003) used both enzymestimulated precipitation and enzyme-labeled antibodymodified electrodes to probe the electron transfer of a redox

pair using electrochemical methods, with the precipitation of an insoluble precipitate on an electrode blocking electron transfer between the redox probe and the electrode. Electrochemical immunosensors that transfer biospecific affinity recongnition reactions into electrochemical signals by biocatalytic precipitation on the electrodes were recently reported (Yoon et al., 2002; Ruan et al., 2002). Avidin, found in chicken egg-white, is a glycoprotein with a molecular weight of 68 kDa. The interaction between avidin (or streptavindin) and biotin exhibits the strongest known affinity (dissociation constant, Kd ∼ 1015 M−1 ) between a ligand and a protein (Weber et al., 1989). The high specificity of avidin for biotin has resulted in the avidin–biotin system being extensively used in biological science and technology, including affinity-based separation, diagnostic assays and DNA hybridization (Cho et al., 1999; Masarik et al., 2003). Avidin is toxic to many organisms due to its ability to deplete biotin, an essential vitamin (Vitamin H), from their environment (Mock et al., 2003). Eating raw egg whites is commonly believed to induce biotin deficiency (Mock et al., 2002). Several methods have been used to analyze for avidin and biotin. These methods include colorimetric competitive inhibition (Gan and Marquardt, 1999), enzymes (Bayer et al., 1986), fluorimetric detection (Reyes et al., 2001), quartz crystal microbalance (Cosnier et al., 2001), and electrochemical detection (Kuramitz et al., 2003; Sugawara et al., 2002). In this work, we describe a bioaffinity system for avidin detection using the magnetoelastic sensor platform. We specifically focus on creating a magnetoelastic sensing surface with the characteristics of specific biological recognition that simultaneously resists non-specific protein adsorption. To achieve this the magnetoelastic sensor is coated with biotinylated poly(ethylene glycol) (PEG). Alkaline phosphatase (AP)-labeled streptavidin and avidin are shown to bind specifically to the biotin-functionalized PEG modified sensor surface, while resisting the non-specific adsorption of streptavidin by the remaining PEG chains. The competitive binding between avidin and a streptavidin–enzyme conjugate for biotinylated PEG immobilized on the sensor surface provides the basis for quantifying avidin. Biocatalytic precipitation of 5-bromo-4-chloro-3-indolyl phosphate through alkaline phosphatase on the magnetoelastic sensor is used to amplify the detected mass change increasing the signal to noise ratio of the detected signal.

2. Experimental 2.1. Chemicals and materials

Fig. 2. Multi-element magnetoelastic sensor arrays defined by laser cutting from a continuous thick-film roll. The tabs spanning the mid-point of the sensors are used to mount the sensor within a fixture for handling.

The ␣-biotin-␻-N-hydroxysuccinimidyl ester of poly (ethylene glycol) carbonate (biotin-PEG-NHS, mol wt ∼ 3.4 kDa), and the N-hydroxysuccinimidyl ester of methoxy-poly(ethylene glycol) propionic (MeO-PEG-SPA, mol wt ∼2 kDa) were obtained from Nektar Therapeu-

C. Ruan et al. / Biosensors and Bioelectronics 19 (2004) 1695–1701

tics, Huntsville, AL. Tris[hydroxymethyl]aminomethane hydrochloride, sodium carbonate, 2-aminoethanethiol hydrochloride (98%), 5-bromo-4-chloro-3-indolyl phosphate disodium salt hydrate (BCIP), methyl alcohol and ethanol were purchased from Sigma-Aldrich, St. Louis, MO. BupHTM phosphate buffered saline pack (0.1 M PBS, pH 7.2) was purchased from Pierce, Rockford, IL, USA. Avidin (hen egg white, 10 mg) and alkaline phosphatase conjugated streptavidin were obtained from Rockland, Gilbertsville, PA, USA. Alkaline phosphatase conjugated streptavidin was diluted 1000-fold to 1 ␮g/ml of streptavidin prior use. Avidin was rehydrated with 1 ml of 0.1 mol/l phosphate buffered saline solution (pH 7.2), and then was serially diluted. Tris buffer (25 mM, pH 9.5) and sodium carbonate buffer (50 mM, pH 9.0) were prepared by dissolving an appropriate amount of tris[hydroxymethyl]aminomethane hydrochloride and sodium carbonate in deionized water, and then adjusting with 1.0 mol/l HCl or 1.0 mol/l NaOH, respectively. Phosphate buffer with 0.01 M sodium phosphate (pH 7.2) was prepared through a 10-fold dilution of the 0.1 mol/l phosphate buffered saline solution. 2.2. Magnetoelastic sensors Magnetoelastic ribbons comprised of Metglas alloy 2826, composition Fe40 Ni40 P14 B6 , were purchased from Honeywell Corporation and used throughout this work. To protect the magnetoelastic sensors from corrosion the sensors were pre-coated with a ≈ 100 nm layer of gold applied by thermal evaporation. Prior to deposition of the gold layers the magnetoelastic sensors were ultrasonically cleaned in water with micro-90 cleaning solution, then in methyl alcohol, and finally in deionized water, with the resulting substrates dried in a stream of nitrogen. The evaporation was done at room temperature at a pressure of approximately 10−7 Torr. After deposition of the protective gold film the sensors were annealed at 200 ◦ C for 3 h in vacuum (∼10−3 Torr) to elim-

1697

inate any residual stress in the sensor, and improve adhesion of the gold layer to the magnetoelastic substrate. The magnetoelastic sensors were defined from a continuous ribbon using computer controlled laser cutting; dimensions of the sensors used in this work are 6.0 mm × 2.0 mm× 28 ␮m. 2.3. Modification of sensors with PEG and biotinylated PEG The gold-coated sensors were immersed in a 10 mM solution of 2-aminoethanethiol hydrochloride in ethanol overnight at room temperature. After reaction, the modified substrates were removed from solution and rinsed several times with ethanol to remove any physically absorbed aminoethanethiols, then dried under a stream of nitrogen. An exposed active amino group formed on the surface of the sensor (Heleg-Shabtai et al., 1997). The sensors were subsequently immersed into 20 mg/ml biotin-PEG-NHS in 50 mM Na2 CO3 , pH 9.0 solution for 2 h at room temperature, rinsed with water, and dried under a stream of nitrogen (Lee et al., 2000). For the PEG modified surfaces, the sensors were immersed into 20 mg/ml MeO-PEG-SA in 50 mM Na2 CO3 , pH 9.0 solution for 2 h at room temperature. To test for non-specific binding the gold coated, gold-surface functionalized with amino group, and the PEG modified sensors were respectively immersed into AP-conjugated streptavidin solution (1:1000) for 1 h at room temperature. The sensors were then washed with 0.01 mol/l PBS buffer solution to remove any loosely adsorbed AP-conjugated streptavidin. 2.4. Assay procedure for avidin Biotin-functionalized magnetoelastic sensors were exposed to 1 ml, 0.01 mol/l pH 7.2 PBS solution containing 1000-fold diluted AP-conjugated streptavidin, and serial

Fig. 3. Box containing microprocessor based electronics used for determining resonance frequency of magnetoelastic sensor, or sensor array, inserted into the coil mounted at the top of the box. A RS232 cable connects the electronics box with a PC providing a user interface and means for data display.

1698

C. Ruan et al. / Biosensors and Bioelectronics 19 (2004) 1695–1701

dilutions of avidin. The sensors were incubated in the test solutions for 1 h at room temperature, approximately 25 ◦ C, after which the sensors were then rinsed with 0.01 mol/l PBS buffer solution to remove any non-specifically bound AP-conjugated streptavidin and avidin.

precipitation of BCIP induces a time-dependent mass change occurring on the magnetoelastic transducer. Thus, the mass sensing capability of magnetoelastic sensor could be exploited for detection, in solution, of alkaline phosphatase immobilized via non-specific adsorption or specific binding between biotin–avidin on the ME sensor surface.

2.5. Signal measurement 3.2. Surface chemistry Microprocessor-based magnetoelastic sensor monitoring electronics employing a frequency counting technique (Zeng et al., 2002) were used to determine the resonance frequency of the different sensors, with a resolution of approximately 2 Hz. The sensors, bound with different amounts of AP-conjugated streptavidin and avidin were put directly into a small vial containing 2 ml of Tris buffer (25 mM, pH 9.5) solution with 2 mg/ml of BCIP. The tube-like vial, containing the test solution and sensor, was inserted into a small coil used to both transmit the interrogation signal and detect the sensor response. Fig. 3 shows the magnetoelastic sensor monitoring electronics box used for this work, in which the signal telemetry coil is mounted at the top; a RS232 cable connects the electronics box to a personal computer providing a software interface. The resultant mass load of dimer of BCIP on the sensor due to precipitation is determined in real time in solution through determination of sensor resonance frequency.

3. Results and discussion 3.1. Biocatalytic amplification on magnetoelasticb sensors The mass sensitivity of a magnetoelastic sensor 6 mm in length is generally equivalent to a 5 MHz quartz crystal microbalance (Ruan et al., 2003; Grimes et al., 2002). The frequency change of a magnetoelastic ribbon, f, resulting from a translational mass load is given by (Schmidt and Grimes, 2001): f = −fr

m 2mo

(1)

where fr is the fundamental resonant frequency of the ribbon-like sensor, and mo is the mass of the sensor. The frequency shift is downward with increasing mass, f = fmass loaded − fr < 0. The fundamental resonant frequency of the longitudinal vibrating uncoated sensor is determined by its length, Young’s modulus E, and density ρ (Schmidt and Grimes, 2001):  E 1 (2) fr = ρ 2L Rather than trying to detect avidin or streptavidin directly we use an enzymatic catalytic amplification process to increase the signal to noise ratio. Alkaline phosphate catalyzed

The well-established self-assembled monolayer (SAM) of organomercaptan molecules on an Au surface provides a simple methodology for fabricating various functionalized surfaces (Wink et al., 1997; Disley et al., 1998). Adsorption of cysteamine (CA) from its ethanol solution on Au surfaces generates an amine functionalized surface that can be used to fabricate chemical and biological sensing surfaces through use of functional binding agents. This is true even though the monolayer of cysteamine on Au is not homogeneous, leaving parts of the surface uncovered (Pradier et al., 2002). Fig. 4a shows the configuration of the avidin or streptavidin recognition layer on the magnetoelastic sensor surface. The gold-coated sensor surface is functionalized with the CA thin film through a self-assembled process. The PEG, or biotinylated PEG, was covalently linked to an amine functionalized sensor through active ester groups. The resulting protein recognition interface is reacted with alkaline phosphatase-labeled streptavidin and avidin to yield a biotin–avidin/streptavidin complex. For interface-based bioassays including biosensors, and especially for those based on competitive inhibition, non-specific adsorption of the analyte competitor limits sensitivity, and often even the feasibility for an assay system due to the noise associated with non-specific adsorption. Previously, Pradier et al. (2002) demonstrated that some extent of non-specific adsorption of avidin was observed by Fourier transform infrared reflection absorption, while directly coupling of NHS-biotin to a monolayer of cysteamine on an Au surface. To decrease this observed non-specific binding, a hydrophilic long-chain amineterminated monolayer of diamine 2,2 -(ethylenedioxy) bis(ethylenediamine) (DADOO) was used to bind biotin molecules. The bifunctional PEG reagent was used in fabricating the magnetoelastic biosensing interfaces (Fig. 4b). Fig. 5 shows the results of non-specific adsorption of alkaline phosphatase conjugated streptavidin on the differently functionalized sensor surfaces. The plot is illustrative of typical results achieved for five sensors each having the specifically functionalized surface coatings. It is clearly demonstrated from Fig. 5a that there is a strong non-specific adsorption of alkaline phosphatase streptavidin conjugates on the gold-coated sensor surface; the resonance frequency of the sensor decreased approximately 220 Hz (0.19% shift in resonance frequency) after 50 min reaction with BCIP in pH 9.5 Tris buffer. This result is consistent with a previous study on adsorption of avidin and streptavidin on Au surfaces (Ebersole et al., 1990), the results of which indicated that a monolayer of avidin and

C. Ruan et al. / Biosensors and Bioelectronics 19 (2004) 1695–1701

1699

O HN

NH

Biotin

O O

O O O

O

O

O

O

O

O

O O

O

O

O

O O

O

O

O

O O

O

O

O

O

O

O

O

O

O

O

O

O O

HN

S

N H

O O

O

O O

O

O

O

O

O

O

O O

HN

O

O

O

O

O O

O O

O

O

O

PEG

HN CA

S

S

S

Au coated magnetoelastic ribbon

(a)

O HN

NH

O CH2-CH2-CH2-CH2 C

S

O

O N H

PEG OCH2-CH2-C O

N O

(b)

Change in Resonant Frequency (Hz)

Fig. 4. (a) Schematic illustration of bifunctional PEG binding reagent. (b) Biomolecular recognition layer for avidin and streptavidin, biotinylated PEG modified magnetoelastic sensor surface.

0

-50

-100

-150

-200 Au coated surface Cysteamine modified surface

-250

-300

PEG functionalized surface

0

10

20

30

40

50

60

Time (Minutes) Fig. 5. The non-specific adsorption characteristics of magnetoelastic sensors. The timedependent frequency response of the sensors with different functionalized surfaces in pH 9.5 Tris buffer containing 2 mg/ml of BCIP. The sensors were immersed in pH 7.2 PBS containing 1000-old diluted alkaline phosphatase streptavidin conjugate for 1 h and washed with PBS buffer before measurement. (a) Freshly evaporated Au surface, (b) cysteamine modified surface and (c) PEG modified sensor surface. The 5% of the measured data points are indicated by symbols.

strepavidin spontaneously formed on gold surfaces. The ellipsometrically determined adsorption layer thickness for avidin and streptavidin was 4.8 and 4.2 nm, respectively. The hydrophobic interactions and ligating functional groups (NH2 , CO2 − ) are the dominant factors in the non-specific adsorption process on gold surfaces. Compared to the gold surface, the extent of non-specific adsorption of streptavidin on the cysteamine surface is considerably decreased, see Fig. 5b. However, there is still some non-specific adsorption of strepavidin observed on the cysteamine-modified Au surface; the resonant frequency of the sensor decreased approximately 50 Hz (0.04% shift in resonant frequency) after 50 min reaction with BCIP in pH 9.5 Tris buffer. Two reasons may account for the non-specific adsorption of strepavidin on the hydrophilic CA modified surfaces. First, it has been reported that the monolayers for short chain thiol compounds on Au surfaces are disordered (Wirde et al., 1999; Kudelski and Hill, 1999). Hence, there are possibly ‘holes’ present in the CA layer, likely due to a disordered monolayer structure, that results in incomplete coverage of the Au surface and where streptavidin molecules are adsorbed. A second possibility is that a disordered CA layer exposes part of the CH2 –CH2 chains to solution, and these hydrophobic carbon–hydrogen chains interact with streptavidin molecules to induce non-specific adsorption.

C. Ruan et al. / Biosensors and Bioelectronics 19 (2004) 1695–1701

0

-50

-100 0 0.4 µg/ml 0.6 µg/ml 0.8 µg/ml

-150

-200 0

10

20

30

40

50

60

Change in Resonant Frequency (Hz)

Change in Resonant Frequency (Hz)

1700

200

150

100

50

0 -0.5

Time (Minutes) Fig. 6. The typical time-dependent frequency response of sensors in the presence of BCIP in pH 9.5 Tris buffer after their incubation in a pH 7.2 PBS solution with different avidin concentrations. The 5% of the measured data points are indicated by symbols.

The PEG modified sensor showed no significant nonspecific adsorption of streptavidin, Fig. 5c; there was no detectable change in resonance frequency during the 50 min reaction with BCIP. This result indicates that immobilized PEG layers can effectively reduce, if not eliminate, non-specific adsorption of proteins on magnetoelastic sensors making the platform well suited for biosensing applications. 3.3. Recognition of avidin and streptavidin The method exploited in this avidin assay is based on the competitive binding between avidin and enzyme streptavidin conjugates for biotinylated PEG immobilized on a magnetoelastic ribbon surface. Sensors reacted with a sample containing a constant amount of enzyme streptavidin conjugates and no avidin were used as the negative controls. Increasing the concentration of avidin in the sample, the amount of enzyme streptavidin conjugates bound to the biotinylated surface decreased proportionally. As indicated, alkaline phosphatase on the magnetoelastic ribbon surface stimulated a surface precipitation reaction with BICP that induced changes in the sensor resonance frequency. Fig. 6 shows the time-dependent response of magnetoelastic sensors to the enzymatic catalytic reaction on the sensor surface with only AP streptavidin conjugates, and with a mixture of avidin and AP streptavidin conjugates. Due to the lack of non-specific adsorption observed on the PEG surface the frequency changes seen in Fig. 6 are those only associated with the specific binding between immobilized biotinylated PEG on the sensor surface and avidin or AP streptavidin in the solution. It is also evident from Fig. 6 that the rate of resonance frequency change, and magnitude of resonance

0

0.5

1

1.5

2

Avidin concentration (µg/ml) Fig. 7. Changes in resonance frequency as a function of avidin concentration in the test solution (␮g/ml).

frequency change, decreases with increasing avidin concentration. Fig. 7 shows the steady-state (50 min) change in sensor resonance frequency versus avidin concentration ranging from 0.2 to 0.75 ␮g/ml. Results indicate changes in resonance frequency are linearly proportional to the avidin concentration. The limit of detection for avidin, defined as the amount required generating a signal of three times the standard deviation of negative controls plus average value of negative control, is 200 ng/ml. The sensitivity of the sensor for avidin is comparable with the colorimetric competitive inhibition method for the quantitation of avidin (Gan and Marquardt, 1999). The detection limit of our sensors is significantly below that obtained with the Bayer microplate procedure (Bayer et al., 1986) of 1000 ng/ml.

4. Conclusions We have demonstrated that the magnetoelastic sensor platform, combined with an enzymatic biocatalytic precipitation system, provides a low cost yet highly sensitive tool for detecting surface events including non-specific adsorption and specific protein binding. Poly(ethylene glycol) used as an interlayer in immobilizing the magnetoelastic sensor surface can effectively prevent non-specific adsorption. Competitive binding between avidin and enzyme labeled streptavidin for surface biotin enable quantification of avidin concentration. Our avidin detection limit is 200 ng/ml. The sensitivity of our magnetoelastic biosensor for avidin is comparable with the colorimetric competitive inhibition method (Gan and Marquardt, 1999). Our study presents a new path to the fabrication of reliable, easy to use biosensors that we hope will be of utility to the scientific community.

C. Ruan et al. / Biosensors and Bioelectronics 19 (2004) 1695–1701

References Barandiaran, J.M., Gutierrez, J., 1997. Magnetoelastic sensors based on soft amorphous magnetic alloys. Sens. Actuators A. 59, 38–42. Bayer, E.A., Ben-Hur, H., Wilchek, M., 1986. A sensitive enzyme assay for biotin, avidin and strepavidin. Anal. Biochem. 154, 367–370. Cai, Q.Y., Grimes, C.A., 2001. A salt-independent pH sensor. Sens. Actuators B 79, 144–149. Chiriac, H., Marinescu, C.S., óvári, T.-A., Neagu, M., 1999. Sensor applications of amorphous glass-covered wires. Sens. Actuators A 76, 208–212. Cho, H.C., Lee, D.J., Kim, S.Y., Kim, J.H., Paeng, I.R., Cha, G.S., 1999. Enzyme-linked competitive binding assays based on the biotin–avidin interaction. Anal. Sci. 15, 343–347. Cosnier, S., Perrot, H., Wessel, R., 2001. Biotinylated polypyrrole modified quartz crystal microbalance for the fast and reagentless determination of avidin concentration. Electroanalysis 13, 971–974. Disley, D.M., Cullen, D.C., You, H.X., Lowe, C.R., 1998. Covalent coupling of immunoglobulin G to self-assembled monolayers as a method for immobilizing the interfacial-recognition layer of a surface plasmon resonance immunosensor. Biosen. Bioelectron. 13, 1213–1225. Ebersole, R.C., Miller, J.A., Moran, J.R., Ward, M.D., 1990. Spontaneously formed functionally active avidin monolayers on metal surfaces: A strategy for immobilizing biological reagents and design of piezoelectric biosensors. J. Am. Chem. Soc. 112, 3239–3241. Ersöz, A., Ball, J.C., Grimes, C.A., Bachas, L.G., 2002. Characterization of electrochemically deposited polypyrrole using magnetoelastic material transduction elements. Anal. Chem. 74, 4050–4053. Gan, Z., Marquardt, R.R., 1999. Colorimetric competitive inhibition method for the quantitation of avidin, strepavidin and biotin. J. Biochem. Biophys. Methods 39, 1–6. Grimes, C.A., Kouzoudis, D., Mungles, C., 2000. Simultaneous measurement of liquid density and viscosity using remote query magnetoelastic sensors. Rev. Sci. Instrum. 71, 3822–3824. Grimes, C.A., Mungle, C.S., Zeng, K., Jain, M.K., Dreschel, W.R., Paulose, M., Ong, K.G., 2002. Wireless magnetoelastic resonance sensors: a critical review. Sensors 2, 289–308. Hanshoff, H., Galla, H.-J., Steinem, C., 2000. Piezoelectric mass-sensing devices as biosensors—an alternative to optical biosensors? Angew. Chem. Int. Ed. 39, 4004–4032. Heleg-Shabtai, V., Katz, E., Willner, I., 1997. Assembly of microperoxidase-11 and Co(II)-protoporphyrin IX reconsitituted Myoglobin monolayers on Au-Electrodes: Integrated bioelectrocatalytic interfaces. J. Am. Chem. Soc. 119, 8121–8122. Jain, M.K., Schmidt, S., Mungle, C., Loiselle, K., Grimes, C.A., 2001. Measurement of temperature and liquid viscosity using magneto-acoustic/magneto-optical sensors. IEEE Trans. Magn. 37, 2767–2769. Katz, E., Willner, I., 2003. Probing biomolecular interactions at conductive and semiconductive surfaces by impedance spectroscopy: Routes to impedimetric immunosensors, DNA-sensors, and enzyme biosensors. Electroanalysis 15, 913–947. Kudelski, A., Hill, W., 1999. Raman study on the structure of cysteamine monolayers on silver. Langmuir 15, 3162–3168. Kuramitz, H., Matsuda, M., Sugawara, K., Tanaka, S., 2003. Electrochemical evaluation of the interaction between avidin and biotin at biotinylated polypyrrole electrode using a redox marker. Electroanalysis 15, 225–229. Lee, G.U., Metzger, S., Natesan, M., Yanavich, C., Dufrˆene, Y.F., 2000. Implementation of force differentiation in the immunoassay. Anal. Biochem. 287, 261–271.

1701

Masarik, M., Kizek, R., Kramer, K.J., Billova, S., Brazdova, M., Vacek, J., Bailey, M., Jelen, F., Howard, J.A., 2003. Application of avidin–biotin technology and adsorptive transfer stripping square-wave voltammetry for detection of DNA hybridization and avidin in transgenic avidin maize. Anal. Chem. 75, 2663–2669. Mock, D.M., Mock, N.I., Stewart, C.W., LaBorde, J.B., Hansen, D.K., 2003. Marginal biotin deficiency is teratogenic in ICR mice. J. Nutr. 133, 2519–2525. Mock, D.M., Henrich, C.L., Carnell, N., Mock, N.T., 2002. Indicators of marginal biotin deficiency and repletion in humans: validation of 3-hydroxyisovaleric acid excretion and a leucine challenge. Am. J. Clin. Nutr. 76, 1061–1068. Puckett, L.G., Barrett, G., Kouzoudis, D., Grimes, C.A., Bachas, L.G., 2003. Monitoring blood coagulation with magnetoelastic sensors. Biosens. Bioelectron. 18, 675–681. Pradier, C.M., Salmain, M., Zheng, L., Jaouen, G., 2002. Specific binding of avidin to biotin immobilized on modified gold surfaces: Fourier transform infrared reflection absorption spectroscopy analysis. Sur. Sci. 502/503, 193–202. Reyes, F.D., Ferna dez Romero, J.M., Luque de Castro, M.D., 2001. Determination of biotin in foodstuffs and pharmaceutical preparations using a biosensing system based on the streptavidinbiotin interaction. Anal. Chim. Acta. 436, 109–117. Rogers, K.R., 2000. Principles of affinty-based biosensor. Mol. Biotechnol. 14, 109–129. Ruan, C., Zeng, K., Varghese, O.K., Grimes, C.A., 2003. Magnetoelastic immunosensors: Amplified mass immunosorbent assay for detection of Escherichia coli O157:H7. Anal. Chem. 75, 6494–6498. Ruan, C., Yang, L., Li, Y., 2002. Immunobiosensor chips for detection of Escherichia coli O157:H7 using electrochemical impedance spectroscopy. Anal. Chem. 74, 4814–4820. Schmidt, S., Grimes, C.A., 2001. Characterization of nano-dimensional thin-film elastic moduli using magnetoelastic sensors. Sens. Actuators A. 94, 189–196. Sugawara, K., Kato, R., Shirotori, T., Kuramitz, H., Tanaka, S., 2002. Voltammetric behavior of avidin–biotin interaction at a biotin/thionine modified Au electrode. J. Electroanal. Chem. 536, 93–96. Su, X.D., O’Shea, S.J., 2001. Determination of monoenzyme-and bienzyme-stimulated precipitation by a cuvette-based surface plasmon resonance instrument. Anal. Biochem. 299, 241–246. Whelan, R.J., Wohland, T., Neumann, L., Huang, B., Kobilka, B.K., Zare, R.N., 2002. Analysis of biomolecular interactions using a miniaturized surface plasmon resonance sensor. Anal. Chem. 74, 4570–4576. Weber, P.C., Ohlendorf, D.H., Wendoloski, J.J., Salemme, F.R., 1989. Structure origins of highaffinity biotin binding to streptavidin. Science 243, 85–88. Wink, T., vanZuilen, S.J., Bult, A., vanBennekom, W.P., 1997. Self-assembled monolayers for biosensors. Analyst 122, R43–R50. Wirde, M., Geluis, U., Nyholm, L., 1999. Self-assembled monolayers of cystamine and cysteamine on gold studied by XPS and Voltammetry. Langmuir 15, 6370–6378. Yoon, H.C., Hong, M.Y., Kim, H.S., 2000. Affinity biosensor for avidin using a double functionalized dendrimer monolayer on a gold electrode. Anal. Biochem. 282, 121–128. Yoon, H.C., Yang, H.S., Kim, Y.T., 2002. Biocatalytic precipitation induced by an affinity reaction on dendrimer-activated surfaces for the electrochemical signaling from immunosensors. Analyst 127, 1082– 1087. Zeng, K.F., Ong, K.G., Mungle, C.M., Grimes, C.A., 2002. Time domain characterization of oscillating sensors: Application of frequency counting to resonance frequency determination. Rev. Sci. Instrus. 73, 4375–4380.