Biosensors

FULL TEXT (TEXT + FULL FIGURES)

Current Opinion in Biotechnology
Vol. 5, No. 1, February 1994
Biosensors
[Review article]
Anthony PF Turner
Current Opinion in Biotechnology 1994, 5:49-53.

Viewing options [Help]

text only

Associated links

Publications by
Anthony PF Turner

Jump to this record in Evaluated MEDLINE

Related records from Evaluated MEDLINE

Related fulltext articles on BioMedNet

Outline

Abstract
Advances in biosensor technology over the past year have included developments in metalized electrodes, mediated electrochemistry, direct electron transfer, impedance measurement, optical immunosensors, optodes, biomimicry, piezoelectric biosensors, enzyme thermistors, in vivo biosensors, surface characterization, organic-phase biosensors and tissue-based biosens- ors. Increasingly, molecular biology and engineering are being used in sensor design.

Abbreviations

FET —field effect transistor;

ISFET —ion- sensitive FET.

Introduction

Biosensors combine the selectivity and sensitivity of biology with the processing power of modern microelectronics to offer powerful new analytical tools with major applications in medicine, environmental diagnostics and the food and drink industries. I have previously published extensive reviews of the area [1][2][3][4][5][6], and the reader is referred to these for a comprehensive historical perspective. The present article updates a previous contribution to this journal [7] by discussing papers published over the past twelve months. A total of 1090 publications that appeared during the year were considered for inclusion, comprising 765 original papers, 179 patents and 146 books or reviews. The selection of a tiny proportion of these for presentation is, of necessity, subjective and is based principally on the perceived contribution of the paper to the solution of key problems in the application of biosensors.

Metalized electrodes

The detection of electroactive products, such as hydrogen peroxide, has proved a very practical approach for the design of enzyme electrodes based on oxidoreductases. Glucose detection using glucose oxidase has received the majority of attention [8]. Rational design at a molecular level has been boosted by the recent elucidation of the three-dimensional structure of this enzyme [9••], but advances in lactate analysis have also been prominent [10]. The relatively high potential required to oxidize hydrogen peroxide at conventional electrodes, however, leads to problems of interference resulting from other electroactive species present in the sample. The most successful commercial solution to this has involved the use of mediators such as ferrocene and its derivatives [11][12]. Recent attention, however, has focused on the alternative strategy of reducing the potential at which hydrogen peroxide oxidation occurs. A 7 µm diameter platinized carbon fibre microelectrode for glutamate has been reported that exhibits a 12 s response time using a pulsed operating potential of 300 mV (versus an Ag/AgCl electrode) [13].

Mediated electrochemistry

Interest in NAD-independent quinoprotein dehydrogenases as components of enzyme electrodes has recently been revived. Two ferrocene derivatives and phenazine methosulphate have been explored as mediators with membrane-bound aldose dehydrogenase from Gluconobacter oxydans [14]. Furthermore, quinoprotein glucose dehydrogenase incorporated in a redox gel has been demonstrated to yield a high current density glucose electrode [15••]. New soluble mediators continue to be elucidated for capillary-fill type devices [16], and much progress has been made in immobilizing mediators to the electrode surface [17], within conducting polymers [18] or to the enzyme itself, in combination with redox polymers [19]. The last of these papers demonstrates the selectivity that can be obtained by electrochemically coupling only one enzyme in a bienzyme electrode to the matrix. On a practical note, Kress-Rogers et al. [20] have recently reported the utility of mediated three-dimensional arrays of electrodes to measure microbial contamination of meat.

Direct electron transfer

The elegance of direct electron transfer between a redox enzyme and an electrode has long been appreciated. If sustained and sizable currents can be achieved, this approach might offer the simplest route for interfacing catalytic proteins to electronic circuits, thus, opening the way for a range of biosensors and other bioelectronic devices. Such an intimate connection would also, presumably, indicate ways to avoid unwanted redox reactions at the electrode and, thus, eliminate the prime source of interference in amperometric biosensors. The direct electrochemistry of electron transfer proteins such as cytochrome c has been extensively studied in recent years. Cooper et al. [21•] have studied the electrochemistry of soluble cytochrome c at a N -acetyl cysteine- modified gold electrode. Using carbodiimide condensation, these authors were able to covalently attach cytochrome c to the electrode, and demonstrate electrochemistry of the immobilized protein. Preliminary results showed a linear calibration for the rate of superoxide production by xanthine oxidase over the range 0– 0.48 µM enzyme, demonstrating the potential of the approach as a sensor for the production of superoxide radicals in vivo. Another study contrasts this approach with alternatives based either on absorption of cytochrome c to a platinized carbon electrode or on low potential detection of hydrogen peroxide from superoxide dismutase at a platinized carbon electrode [22]. It was concluded that cytochrome c absorbed to platinized carbon demonstrated the greatest sensitivity, and this method was subsequently used to measure free radical production by stimulated human neutrophils.

Although direct electrochemistry of electron transfer proteins such as cytochrome c is now relatively well established, few reports have appeared describing electron transfer between electrodes and the larger redox enzymes. Where observed, electrochemistry has been typically poor as a result of the existence of a thick insulating protein shell around the active centre of the enzyme. In a radical new approach to the construction of redox-enzyme electrodes, Koopal et al. [23••] immobilized glucose oxidase on a conducting polymer located within the pores of a track-etch membrane. Polyconjugated conducting materials such as polypyrrole and polythiophene can form microtubules in the pores of such membranes. These authors postulated that such structures can communicate directly with the redox centre of glucose oxidase. A highly selective amperometric sensor, which is apparently unaffected by oxygen concentration, was constructed that measured glucose in the range 1–30 mM. The argument was presented that direct electron transfer was occurring. The same group have expanded this concept in a later publication [24], which also addresses the problem of the difficulty of manufacturing track-etched membranes. They propose the use of uniform latex particles as a porous matrix for polypyrrole coating and subsequent enzyme immobilization. This will facilitate the mass production of sensors [25] because the latex particles employed can be formulated as an ink and printed. A similar mechanism of interaction has been reported for polyethylene glycol modified glucose oxidase in a carbon paste electrode [26], although the apparent current density is considerably lower.

Impedance measurement

Practical 'immuno' field effect transistors (immunoFETs) have proved difficult to realize as a result of, among other things, screening of the protein charge by small counterions from the electrolyte. Kruise et al. [27•] have proposed that protein charge on an ion-sensitive field effect transitor (ISFET) surface can be measured using a.c. impedance. Using lysozyme membranes, they demonstrate that measurement of protein charge density changes, in response to pH shifts, is feasible.

Optical biosensors

The measurement of refractive index changes associated with biological affinity reactions, such as antibody–antigen binding, offers exciting possibilities for unlabelled direct immunoassay. Pharmacia Biotech GmbH already have a successful instrument on the market based on surface plasmon resonance ([28]; this issue, pp 65–71), and other commercial devices based on related principles are expected. Clerc and Lukosz [29] have described an integrated optical output grating coupler biosensor capable of direct detection of 3 × 10 ^{- 9} M bovine serum albumin. Sensitivity is one of the principle hurdles to widespread application of such techniques. Using a Mach–Zehnder interferometer on a planar waveguide, a 40 kDa protein was measured to concentrations as low as 5 × 10 ^{- 11} M [30]. Schlatter et al. [31••] at Hoffmann-La Roche have achieved similar limits of detection for IgG using a two-mode interferometric thin-film optical waveguide sensor. Hepatitis B surface antigen was detected at a concentration of 2 × 10 ^{- 13} M in undiluted human serum.

Optical alternatives to enzyme electrodes continue to be developed. A report from one of the leading groups in this area describes an optode for lysine that is based on lysine decarboxylase and a caverine-sensitive membrane [32•]. This membrane consists of plasticized PVC, which contains a lipophilic tartrate as an amine carrier. The transport of cadavarine is coupled to transport of a proton, which causes a spectral change in an indicator dye.

Piezoelectric and thermal biosensors

The successful application of piezoelectric materials in biosensors has been limited because of both the lack of supporting theory when devices are immersed in liquids and the generally poor sensitivity in their main application as affinity sensors. Walton et al. [33] describe a possible solution to the sensitivity problem by using a very thin polymer film as an acoustic medium. Changes in mass bound to the surface of the film alter the transit time of acoustic waves propagating through the film. The low mass of the piezoelectric layer resulted in a 30-fold enhancement in sensitivity.

Few reports have appeared describing true biosensors using thermal transducers. However, work on the enzyme thermistor (a bioreactor with differential thermal sensing) continues, with an on- site industrial application being described by the authors who originated the technique [34].

In vivo biosensors

In vivo sensing was one of the main goals of Professor Leland C Clark Jr when he first described the concept of the biosensor 31 years ago. Commercial realization of such a device is still awaited, but important advances have recently been reported. Eli Lilly & Co recently presented a long-awaited report on its major programme for developing subcutaneous glucose sensors [35••]. This report describes the mass production of these biosensors and their evaluation in normal human subjects. Unfortunately, certain design problems still need to be ironed out before the device can be commercialized. Urban et al. [36] have described a bienzyme glucose biosensor suitable for in vivo application that is integrated with a pH sensor, which is fabricated on a flexible polyimide base. The device was evaluated in serum and blood.

One of a series of papers on a promising subcutaneous glucose sensor and associated instrumentation has recently been presented by Moatti-Sirat et al. [37•]; this instrument is covered by a recently published World Patent Application WO91/15993. The use of microdialysis as an alternative to implanting the sensor itself is described by Mascini's group [38] and is being developed by an Italian company. If non-invasive monitoring of glucose and other analytes could be performed with simple and inexpensive instrumentation, this would probably supersede the use of chemical sensors (M Arnold, abstract, Am Chem Soc, 1992, 204:99).

Molecular engineering

The importance of surface characteristics in determining the performance of a biosensor is increasingly being recognised. Scanning tunnelling microscopy has proved a particularly useful tool in studying enzyme electrodes [39].

The use of synzymes, artificial enzymes or other analogues of biological behaviour offers a powerful conceptual approach to the design of stable sensors. The late Professor Simon and colleagues presented an elegant comparison of optodes based on alcohol dehydrogenase and a lipophilic amide of trifluoroacetylaniline as a chemical alternative [40••]. The artificial alcohol ligand lacked metabolic activity but offered interesting opportunities for the design of new sensors for the food and drink industry. Another aspect of biomimicry is represented by further work on the artificial nose, which employs chemical sensor arrays in conjunction with artificial neural networks [41].

The concept of using organic phases to enhance the performance of biological elements used in the construction of biosensors has been vigorously pursued and extended to include tissue as the biological catalyst by Professor Wang's group in New Mexico [42]. Phenolic and peroxide species were detected in chloroform using mushroom, banana or horseradish root as the biocatalyst. The same group have recently published the first description of organic-phase biosensing of enzyme inhibitors [43••].

Professor Rechnitz, the first researchers to use plant-based biosensors, has extended their lifetime by substituting plant tissue grown in aseptic culture [44]. Tobacco callus tissue sensors for the determination of hydrogen peroxide are reported to have a lifetime of over four months.

Aizawa's group has recently presented two exciting papers illustrating the potential of molecular biology in biosensor design. A fusion protein combining protein A and firefly luciferase has been produced and used in a bioluminescent immunoassay for human IgG [45]. In a second publication, firefly luciferase is fused with the TOL plasmid responsible for the degradation of benzene and its derivatives to produce a luminescent Escherichia coli strain, which can be used to monitor environmental pollution [46••].

Conclusions

The potential applications for biosensors (and the hurdles to their introduction) have been thoroughly documented in the literature. Biosensor technology is, by definition, an applied field, and its success or failure in the forthcoming decade will depend on the demonstrable solution of significant analytical problems. Emerging information about molecular structures and mechanisms are providing the basis for rational and systematic engineering of biosensors to meet clearly defined needs.

Acknowledgement

Many thanks to Professor Mascini and the University of Florence for inviting me to take up the visiting chair, which allowed me to update my reading in preparation for this review.

References and recommended reading
Papers of particular interest, published within the annual period of review, have been highlighted as:
• of special interest.
•• of outstanding interest.

Turner APF, Karube I, Wilson GS: Biosensors: Fundamentals and Applications. Oxford: Oxford University Press, 1989,