IVD Technology Magazine
IVDT Article Index

Originally published September, 1997

Electrochemical biosensors for affinity assays

Part 2

Mark S. Vreeke

To develop commercially successful affinity biosensors, IVD firms will first need to identify niche markets suited to this technology.

Note: this is the second part of a two-part article. Part 1 is also available for on-line reading.

In recent years, glucose monitors using biosensor technologies have enjoyed tremendous commercial success. Despite their seeming potential for use in other areas, however, biosensors for other analytes have so far met with limited endorsement. Advancing the use of this emerging technology will require manufacturers to have an understanding of both the opportunities and limitations it presents.

The first installment of this article provided a brief introduction to biosensors, with emphasis on amperometric enzyme electrodes (IVDT, July/August 1997, pp 39­45). Amperometric biosensors combine the selectivity of an enzyme reaction with the sensitivity of amperometric detection. In operation, these biosensors use an enzyme to convert an analyte into an electroactive product, which is then transduced into a quantifiable amperometric response by an electrode.

The level of sophistication associated with such biosensors can be defined by the manner in which the enzyme reaction is transduced to the amperometric response. The latest generation of biosensors is characterized by "wired" enzymes, in which the enzymatic reaction is directly transduced to the amperometric response by means of a molecular wire that connects the enzyme to the electrode.

Below, the second installment of this article describes the use of wired-enzyme technology with peroxidase enzymes to detect H^₂O^₂. Although H^₂O^₂ is rarely an analyte of primary interest, diagnostic assays often require H^₂O^₂ detection. Equally important, the response characteristics of H^₂O^₂ biosensors can facilitate several unique applications, including the adaptation of wired-enzyme sensors to an electrochemical affinity assay.

Wiring of Peroxidase

Peroxidases (POD) include a broad group of enzymes able to catalyze the following reactions:

The first reaction is highly selective for peroxides, primarily H^₂O^₂ and a few small organic peroxides. The second and third reactions are much less selective for electron donors (HA).

Amperometric peroxidase-based H^₂O^₂ sensors have been made by using fast reversible redox couples. In these, the reducing member of the redox couple (essentially species HA in the reactions above) donates electrons to H^₂O^₂ and is oxidized:

The oxidized redox couple is then cathodically reduced at the electrode surface:

The most commonly used enzyme in these biosensors is horseradish peroxidase (HRP), a small (44-kD) heme peroxidase, but others have been used as well. Detection schemes vary in their method of enzyme immobilization, mediator, and type of enzyme. At one extreme are systems based on the direct transfer of electrons from the electrode surface through surface-bound mediators to HRP redox centers contacting the surface. At the other extreme are systems with freely diffusing mediators and enzyme.

The redox polymers designed and used for oxidase wiring are able to transfer redox equivalents in the reverse direction from electrode to POD active sites, making wired-enzyme H^₂O^₂ biosensors (Figure 1).^1,2 When the electrode is poised at 0.0 mV versus SCE, the H^₂O^₂ flux is measured as a cathodic current. The sensitivity is 1 A cm^­2 M^­1 over the linear range extending from 0.1 to 100 µM. The limit of detection for a steady-state measurement is 10 nM. Using a flow system, injections of less than 100 pM have been detected.³

Figure 1. The redox cycles occurring at a three-dimensional redox epoxy wired enzyme electrode. The wired enzyme is a heme-containing peroxidase (POD).

Amperometric biosensors have used platinum electrodes for H^₂O^₂ detection since they were first developed. The platinum electrodes continue to be used because of their excellent performance. Although feasible, POD-based amperometric H^₂O^₂ detection is not commercially used except in a flow-injection analysis detector sold by Bioanalytical Systems (West Lafayette, IN).

With the development of the wired POD biosensor for H^₂O^₂ detection, the supremacy of H^₂O^₂ detection on platinum was challenged. The H^₂O^₂ biosensor also facilitated several applications achievable only due to the enhanced performance or unique characteristics of enzyme wiring. These include sensitive flow-system detectors,^3,4 selective-scanning electrochemical probe tips,^5,6 enhanced sensitivity for oxidase sensors,⁷ amperometric NADH detection,¹ and use in affinity sensors.^8,9

Electrochemical Affinity Biosensors

Affinity sensors detect the coupling reaction between the selective binding unit (SBU) (e.g., avidin, antibody, single-stranded DNA, lectin, and host artificial molecular recognition species) and its complementary component (e.g., biotin, antigen, complementary single-stranded DNA, sugar sequence, and guest target compound). Affinity sensors' design ensures that binding of SBU and complement takes place on the transducer surface. The sensors are thus implicitly heterogeneous. The transducer converts the binding event into a measurable response.

The sensors can be divided into two categories: nonlabeled and labeled. Nonlabeled affinity sensors directly detect the affinity complex by measuring physical changes at the transducer induced by complex formation. In contrast, labeled affinity sensors incorporate a sensitively detectable label, and the presence of the affinity complex is then determined through measurement of the label.

Typically, detection of the binding event is not a direct measurement. Labeling of either SBU or complement aids in signaling the binding event. Enzyme labels are particularly useful for providing signal amplification. Their incorporation yields higher sensitivity. Since enzyme electrodes effectively coupled redox enzymes with amperometric detection, there was a natural progression to coupling enzyme-labeled affinity reactions and amperometric detection.

Heineman pioneered the use of alkaline phosphatase­antibody conjugates to perform sandwich immunoassays.¹⁰ In these assays, aminophenyl phosphate is used as a substrate (in place of nitrophenyl phosphate), and the aminophenol product is detected anodically with a flow-injection analysis system.

Aizawa devised a host of sensors with a classic Clark-type O^₂ electrode as the base sensor and catalase as the enzyme label.¹¹ Catalase is used to decompose H^₂O^₂ to O^₂ and H^₂O. When the enzyme label is immobilized at the sensor surface by an affinity reaction, an increase in O^₂ signal is observed in an H^₂O^₂ solution.

Rishpon, Bourdillon, and others have developed electrode-based affinity sensors incorporating enzyme labels and the immobilization of an affinity component at the electrode surface.^12­15 Although excellent sensitivities were obtained, these assays were hampered by the need to wash the working electrode and change the incubation and test solutions. The chief problem was the difficulty in distinguishing enzyme-catalyzed reactions in bulk solution from surface-associated reactions.

A goal of affinity sensor engineers has been the development of nonseparation methods where wash steps (a source of irreproducibility) are not necessary. In a recent article, Duan and Meyerhoff proposed a scheme where the substrate, which is converted into electroactive product, is brought into the cell from behind the electrode.¹⁶ This approach allowed for measurement of the binding reaction without the usual washing steps. However, a specially designed cell and electrodes were necessary.

Wired-Enzyme Affinity Biosensors

Previously, sensitive H^₂O^₂ electrodes built by covalently immobilizing HRP in a redox hydrogel were described.^1,2 The redox hydrogel was formed of HRP and water-soluble poly(vinylpyridine) that was quaternized partly with 2-bromoethylamine and partly with osmium bipyridine redox centers (PVP-NH^₂-Os), and cross-linked with poly(ethylene glycol diglycidyl ether) on vitreous carbon.

The sensitivity of these electrodes, on which H^₂O^₂ was electrocatalytically reduced by the sequence shown in Figure 1, was remarkably high: 1 A cm^­2 M^­1. Catalytic electroreduction of H^₂O^₂ was observed with as little as 1 µg/cm² HRP incorporated in the hydrogel. Modification of the catalytic behavior of the hydrogel by addition of minute amounts of HRP led to the hypothesis that the specific binding of HRP-labeled affinity reagents to an electrode could be selectively detected and that the resulting amperometric affinity sensors would not require washings or separation of reagents.

Based on this hypothesis, fast, compact, inexpensive, and separation-free amperometric affinity sensors for biotin and avidin were developed.⁸ The sensor was constructed by immobilizing an SBU (avidin) into a three-dimensional electron-conducting redox hydrogel (enzyme wiring) on a 3-mm vitreous carbon electrode. The SBU provided the electrode affinity for the SBU's complementary component (biotin). Incubation of the affinity sensor with its complementary component led to selective uptake of the complement from the solution. If the complement was first labeled with a redox enzyme, incubation led to binding of that enzyme to the wiring gel on the electrode. The principles of detection in such a sensor for the avidin-biotin system are presented schematically in Figure 2.

Figure 2. Direct transduction of biotinylated-HRP (B-HRP), avidin (X), and biotin (*) concentrations to currents in a PVI-Os "wire" and avidin-modified electrode. When B-HRP binds with avidin in the HRP-wiring hydrogel (A), a current flows. The current is inhibited if (B) the B-HRP binds to dissolved rather than surface avidin or (C) the binding sites of avidin in the wiring hydrogel are occupied by free biotin.

The biotin/avidin affinity electrode was used to directly detect redox enzyme­labeled complement in a test sample (pictorially described in Figure 2, path A). Here the avidin is immobilized at the electrode in the hydrogel, and the conjugate biotin is labeled with a POD redox enzyme. In this method, the affinity sensor is incubated in a solution containing redox-labeled complement (B-HRP). Binding of the complement selectively immobilizes redox enzyme in the hydrogel. Addition of redox enzyme substrate generates an electrical signal detectable at the electrode.

In the case of POD labels, electrons generated at the electrode are relayed to the POD enzyme through the hydrogel network to which the POD is selectively bound by avidin. In the presence of the enzyme's substrate H^₂O^₂, the electrons are then transferred from the reduced POD to hydrogen peroxide, generating the flow of an electrical current. This current is a function of the concentration of biotinylated peroxidase immobilized at the electrode by the SBU. As shown in Figure 3, electrons are relayed from the electrode through the wire and the POD enzyme to H^₂O^₂, which is electroreduced to water. Measurement is generally at +100 mV (Ag/AgCl). PVI-Os is a polymer with a polyvinylimidazole backbone and osmium bipyridine redox sites coordinated to 20% of the imidazole groups. PVI-Os serves the same redox wiring function as PVP-Os-NH^₂.

Figure 3. Time dependence of the current of the polyvinylimidazole with coordinated osmium bipyridine (PVI-Os) avidin-modified electrode (2 µg avidin, 3.3 µg PVI-Os, and 0.83 µg polyethylene glycol diglycidyl ether [PEGDGE]) after injecting H^₂O^₂ to 100 µM and injecting B-HRP to 1 µg/ml concentration. Conditions: 5 ml PBS; 1000 rpm; +0.1 V Ag/AgCl.

The affinity electrode can also detect SBU by a competitive process. An unknown concentration of avidin, free in the solution, is allowed to compete with electrode-immobilized avidin for a limited number of enzyme-labeled complement molecules. This process is pictorially represented in Figure 2, path B. The free avidin effectively prevents the complement from complexing with the avidin in the wiring hydrogel. The current resulting from the electrocatalytic reduction of H^₂O^₂ is higher when fewer complement SBUs are present in the solution. The limit of detection is below 5 µg/ml.

In a similar assay, biotin was detected by allowing a fixed number of labeled complement molecules to compete with an unknown concentration of biotin (not redox enzyme­labeled) for the limited number of SBUs immobilized at the electrode. The process is pictorially presented in Figure 2, path C. The current generated from reduction of enzyme substrate is inversely related to the amount of unlabeled complement. The limit of detection is below 10 nM. All of these assays were accomplished without washing of the electrodes.

As yet, no wash solution has been found that effectively separates biotin from avidin without destroying the ability of avidin to bind biotin or changing the redox characteristics of the PVI-Os films. Such a solution is bound to be elusive, considering that the couple does not separate even at extremes in pH.

The lack of reversibility makes it necessary to use multiple electrodes when establishing calibration curves. Preliminary work with an antibody to biotin incorporated in PVI-Os gels on electrodes has shown that, like the PVI-Os-avidin films, the binding of B-HRP can be tracked by the increase in H^₂O^₂ reduction current. However, unlike the PVI-Os-avidin films, where binding is practically irreversible, the B-HRP binds reversibly to the antibiotin-containing film. In three cycles of binding and separation, the current increased and decreased reproducibly, showing that the film did not degrade upon brief cycling (Figure 4). With any multiple-use affinity biosensor, a washing sequence is required, at least for the separation and removal of the initially bound complement.

Figure 4. Three biotin-labeled horseradish peroxidase binding (B-HRP) cycles (A, B, and C) are shown for an immunosensor made with a 1-µl loading of solution containing 2.5 mg/ml PEGDGE, 1 mg/ml goat antibiotin, and 10 mg/ml PVI-Os mixed in a 1:5:1 ratio. The B-HRP binding event was carried out in 5 mL pH 7.4 PBS. The H^₂O^₂ concentration was 0.1 mM and the B-HRP concentration was 1 µg/ml. The electrode was rotated at 1000 rpm and poised at 100 mV versus Ag/AgCl. The binding was reversed by washing the electrode in pH 2 PBS for 2 hours.

Conclusion

This previous work described a generic approach for direct electrical detection of the occurrence of an affinity reaction. The sensitivity and detection limits were adequate for some widely performed assays. The microampere currents measured were a thousandfold higher than those routinely measured with simple and inexpensive ($50) potentiostats. They were a millionfold higher than currents measured in Faraday cages with state-of-the-art low-noise current amplifiers and potentiostats.

While sensitivity in a competitive assay is typically based on the shape of the displacement curve, the electrochemical assays were actually limited by the electrodes' size and binding capacity. Considering that all the affinity reagent was stripped from a large (5-ml) volume, no obstacle can be seen to detecting thousandfold and even millionfold smaller amounts of affinity reagents, simply by using smaller electrodes. For example, by using standard 10-µm-diameter ultramicroelectrodes, the sensitivity could be increased by a factor of 10⁵.

Over the past decade, biosensors have been touted as the future of chemical sensors. Academic and basic research efforts have flourished. At a recent symposium on biosensing and biosensors sponsored by the American Chemical Society, 250 papers were presented. However, with the exception of blood glucose monitoring, the gap between R&D and development of actual commercial products has rarely been bridged.

The nature of the biosensor limits the opportunities for commercial success. Affinity biosensors will have a difficult time competing with techniques such as standard enzyme-linked immunosorbent assays, which can be fully automated and operated in multiplexed batches of 96 and even 384 samples. Biosensors seem best suited for limited-use and point-of-care applications.

The personal blood glucose­monitoring business is the prime example of a market requiring immediate on-site analysis without requiring high throughputs. To successfully commercialize affinity biosensors, a similar niche market will have to be identified. Likely targets include infectious disease detection, military applications for immediate detection of hazardous chemicals/microbes, food safety monitoring for bacteria, and possibly genome testing.

One hurdle to tackling the limited markets is the difficulty in recovering product development costs. The blood glucose market is several billion dollars strong and can sustain major R&D efforts. The market for an affinity biosensor is only a fraction of this market. A biosensor strategy that is adaptable to multiple analytes will have the distinct advantage of spreading development costs over several products.

Glossary

amperometry: Measurement of the current resulting from a redox reaction.

anode: Electrode at which oxidation occurs.

avidin: A glycoprotein having four subunits. Each subunit has one binding site for biotin.

biotin: A 244-molecular-weight vitamin found in tissue and blood. It binds with a high affinity to avidin.

cathode: Electrode at which reduction occurs.

mediator: Any chemical species able to transfer electrons between an enzyme's active site and an electrode.

oxidation: A redox reaction involving the loss of electrons.

oxidoreductase: An enzyme that catalyzes an electron transfer reaction.

potential (electrochemical potential): The tendency of a species to give off (oxidize) or take up (reduce) electrons. The value is always relative to another reaction.

reduction: A redox reaction involving the addition of electrons.

selective binding unit (SBU): The biological recognition element in the affinity biosensor. It may be an antibody, antigen, DNA sequence, lectin, avidin, or biotin.

References

1. Vreeke MS, Maidan R, and Heller A, "Hydrogen Peroxide and Beta-Nicotinamide Adenine Dinucleotide Sensing Amperometric Electrodes Based on Electrical Connection of Horseradish Peroxidase Redox Centers to Electrodes through a Three-Dimensional Electron Relaying Polymer Network," Anal Chem, 64:3084­3090, 1992.

2. Vreeke MS, Yong KT, and Heller A, "A Thermostable Biosensor of Hydrogen Peroxide," Anal Chem, 67:4247­4249, 1995.

3. Yang L, Janle E, Huang T, et al., "Application of 'Wired' Peroxidase Electrodes for Peroxide Determination in Liquid Chromatography Coupled to Oxidase Immobilized Enzyme Reactors," Anal Chem, 67:1326­1330, 1995.

4. Huang T, Yang L, Gitzen J, et al., "Detection of Basal Acetylcholine in Rat Brain Microdialysate," Chromatog B, 670:323­327, 1996.

5. Horrocks BR, Schmidtke D, Heller A, et al., "Scanning Electrochemical Ultramicroelectrodes for the Measurement of Hydrogen Peroxide at Surfaces," Anal Chem, 65:3605­3614, 1993.

6. Sakai H, Baba R, Hashimoto K, et al., "Local Detection of Photoelectrochemically Produced H^₂O^₂ with a 'Wired' Horseradish Peroxidase Microsensor," J Phys Chem, 99:11896­11900, 1995.

7. Ohara TJ, Vreeke MS, Battaglini F, et al., "Bioenzyme Sensors Based on Electrically Wired Peroxidase," Electroanal, 5:825­831, 1993.

8. Vreeke MS, Rocca P, and Heller A, "Direct Electrical Detection of Dissolved Biotinylated Horseradish Peroxidase, Biotin, and Avidin," Anal Chem, 67:303­306, 1995.

9. Vreeke MS, and Rocca P, "Biosensors Based on Cross-Linking of Biotinylated Glucose Oxidase by Avidin," Electroanal, 8:55­60, 1996.

10. Xu Y, Halsall HB, and Heineman WR, "Heterogeneous Enzyme Immunoassay of Alpha-fetoprotein in Maternal Serum by Flow Injection Amperometric Detection of 4-aminophenol," Clin Chem, 36:1941­1944, 1990.

11. Aizawa M, "Enzyme-Linked Immunosorbent Assays Using Oxygen-Sensing Electrode," in Electrochem Sensors Immunolog Anal, Ngo TT (ed), New York, Plenum, pp 269­278, 1987.

12. Hadas E, Soussan L, Margalit IR, et al., "A Rapid Sensitive Heterogeneous Immunoelectrochemical Assay Using Disposable Electrode," J Immunoassay, 13:231­252, 1992.

13. Bourdillon C, Demaille C, Gueris J, et al., "A Fully Active Monolayer Enzyme Electrode Derivatized by Antibody Attachment," J Am Chem Soc, 115:12264­12269, 1993.

14. Gleria KD, Hill HAO, McNeil CJ, et al., "Homogeneous Ferrocene-Mediated Amperometric Immunoassay," Anal Chem, 58:1203­1205, 1986.

15. Willner I, Blonder R, and Dagan A, "Application of Photoisomerizable Antigenic Monolayer Electrodes as Reversible Amperometric Immunosensors," J Am Chem Soc, 116:9365­9366, 1994.

16. Duan C, and Meyerhoff ME, "Separation-Free Sandwich Enzyme Immunoassay Using Microporous Gold Electrodes and Self-Assembled Monolayer/Immobilized Capture Antibodies," Anal Chem, 66:1369­1377, 1994.

Mark S. Vreeke, PhD, is a product development scientist at TheraSense, Inc. (Alameda, CA). This work was completed at the Department of Chemical Engineering and Materials Science and Engineering Center of the University of Texas at Austin. Support was provided by an H. H. Dow Memorial Award, a Welch Fellowship, NSF, NIH, and the Department of Defense.

Return to Part 1 of this article

Return to IVDT home page.