Definition and Classification

The repertoire of chemicals that can be determined by the sensors mentioned here is relatively limited. To determine the presence or concentration of more complex biomolecules, viruses, bacteria, and parasites rn vivo, it is necessary to borrow from nature (Fig. 7). Biosensors are sensors that use biological molecules, tissues, organisms, or principles. This definition is broad and by no means universally accepted, although it is more restrictive than the other common interpretation that would include all the sensors described in this chapter. The leading biological components of biosensors are summarized

In ermine MEDICAL

DM far *oph{*tlact*d nnftar*

In ermine MEDICAL

DM far *oph{*tlact*d nnftar*


FIG. 7. "lllc development ot biosensors is driven by increased need tor biochemical information in the medical community, the knowledge that nature senses these chemicals best, combined Through emerging technologies to interface the hiochemicals with physical transducers


FIG. 7. "lllc development ot biosensors is driven by increased need tor biochemical information in the medical community, the knowledge that nature senses these chemicals best, combined Through emerging technologies to interface the hiochemicals with physical transducers working chemistry

FIC. 6. A schematic drawing of probe ot the Cardiovascular Devices Svsiem 1000 fiber optic blood gas sensor, based on three diherenl combinations of selective membranes and fluorescein probes. Liehi enters and leaves trom rhe left thermocouple up coating thermocouple up coating working chemistry

TABLE 4 Biological Components of Biosensors"



Antibodies Nucleic acids Receptor proteins Small molecules lonophores

Enzymes Organelles Tissue slices Whole organisms

"The two categories are not mutually exclusive, because, for example, some enzymes may be employed for binding alone.

in Table 4. Enormous progress has been made in the development of biosensors in recent years, and this work has been recently and exhustively reviewed (Turner et al.., 1987). Most of the applications have been in the realm of analytical chemistry for use in chemical processing and fermentation, with the exception of the development of enzyme-based glucose sensors, on which we will focus.

Currently, biosensors are commercially available for glucose (used first in an automated clinical chemistry analyzer and based on glucose oxidase), lactate, alcohol, sucrose, galactose, uric acid, alpha amylase, choline, and L-lysine. All are ampero-metric sensors based on 02 consumption or H202 production in conjunction with the turnover of an enzyme in the presence of substrate. A urea sensor is based on urease immobilized on a pH glass electrode (Turner, 1989). Most of these sensors are macroscopic and are employed in the controlled environment of a clnical chemistry analyzer, but the ExacTech device, manufactured by Baxter since 1987, is a complete glucose sensor containing disposable glucose oxidase-based electrodes, power supply, electronics, and readout in a housing the size of a ballpoint pen. One places a drop of blood on the disposable electrode and a few seconds later a fairly accurate reading of blood glucose is obtained. It is widely believed that much more frequent measurement of blood glucose with correspondingly frequent adjustments of the dose of insulin delivered could significantly improve the long-term prognosis for insulin-dependent diabetics. Increasing the frequency of the current sampling method (i.e, puncturing the finger for drops of blood) is not acceptable. Much progress has been made toward the goal of producing a glucose sensor that could be implanted for a period of time in the tissue or blood, but the problem is a formidable one that epitomizes the attempt to develop biosensors for in vivo use.


The Utility of Biochemical Approaches to Sensing Some of the many advantages of using biochemicals for sensing are summarized in Table 5. The most important is that, despite the disadvantage of using chemically labile components in a sensor, they allow measurement of chemical species that cannot otherwise be sensed. Sensors have been fabricated that incorpo rate small biochemicals such as antibodies, enzymes and other proteins, ion channels, liposomes, whole bacteria and eukaryo-tic cells (both alive and dead), and even plant and animal tissue.

Immobilization One of the key engineering problems in biosensors is the immobilization of the biochemistry used to the transducing device. Approaches range from simply trapping an enzyme solution between a semipermeable membrane and a metal electrode, to covalently cross-linking several enzymes to a porous hydrogel coated on a pH electrode, to covalently cross-linking a complete monolayer of antibody to the surface of an optical fiber. The immobilization of a layer of material over the transducing device increases the response time, so for altered sensitivity and greatly enhanced selectivity, speed is often sacrificed. Monolayers do not contain much material, so to detect binding of so few molecules, it is generally necessary to employ some amplification scheme, such as attachment of an enzyme to an antibody that announces its presence by converting a subsequently added substrate to a large quantity of readily detected product. Such schemes add complexity and time to the detection. Immobilization also has unpredictable effects on the activity and stability of biochemicals.

TABLE 5 Use of Biochemicals for Chemical Detection

Advantages for binding "Uniquely" high selectivity

Possibility of raising antibodies to nearly all antigens Antibodies and biotin-avidin system allow selective attachment of markers and reporters of binding High binding constants possible Several possible detection modalities Ion flux through gated channels can provide gain

Advantages for catalysis For every biochemical there is an enzyme that can be used to detect its presence High selectivity possible with some enzymes Several possible detection modalities Enzymatic cascades can provide gain

Universality of redox coupling and pH effects permit common transduction schemes

Disadvantages of biosensors

Biomolecules generally have poor thermal and chemical stability compared with inorganic materials The function of the biological component usually dictates that they must have narrow operating ranges in temperature, pH, ionic strength Susceptibility to enzymatic degradation is universal Need for bacteriostatic techniques in their fabrication Time-dependent degradation of performance is guaranteed with the use of proteins Production and purification can be difficult and costly Immobilization can reduce apparent activity of enzymes or kill them outright Most live organisms need care and feeding

Sensing Modalities

Potential-Based Sensors (pH and ISE) Some of the first biosensors employed enzyme-catalyzed reactions (such as those of penicillinase, urease, and even glucose oxidase) that affect pH, By putting a pH electrode into the solution containing the enzyme, it is possible to monitor the rate of enzymatic turnover. It is also possible to use pH to monitor the change in production of C02 by bacteria in the presence of substrates that they are capable of metabolizing (Simpson and Kobos, 1982). However, there is always a problem for in vivo use of pH-based sensors related to the fact that the external environment is capable of strong buffering of pH changes, and any change in pH in the immediate environment of the sensing surface is reduced toward the bulk pH by a degree that depends on the strength of that buffering.

Electrochemical Sensors Many enzymes perform oxidation and reduction reactions and can be coupled, if indirectly, to electrodes. The electrochemically active species in enzymes is generally a cofactor (Table 3) that, when bound, is not accessible to the electrode surface at which the electron transfer must take place for detection. In the case of the glucose oxidase reaction, the normal biological reaction is:

The enzyme uses a flavin adenine dinucleotide (FAD) coenzyme to mediate the oxidation, and the resultant FADH2 is directly oxidized by 02 to return to FAD to prepare for the next catalytic reaction. Unlike nicotinamide adenine dinucleotide (NAD) and NAD phosphate (NADP), FAD is tightly bound to the enzyme, so normally only a small diffusible molecule like 02 can gain access to it to alter its oxidation state. This means that under many circumstances, such as those present in tissue, the concentration of 02 is rate limiting, so the sensor often measures not glucose but the rate at which 02 is rate limiting, so the sensor often measures not glucose but the rate at which 02 can arrive at the enzyme to reoxidize its cofactor. There are two electrochemical ways to couple the reaction to electrodes: monitoring depletion of 02 by reducing what is left at an electrode, or monitoring buildup of H202 by oxidizing it to Q2 and protons. The latter approach is generally used to avoid direct effects of Oz variation on the electrode, but this does not completely solve the problem. The electrode reaction for peroxide oxidation is as follows:

The best solution to date to cast off the tyranny of the rate-limiting step of O , diffusion has been the use of electrochemical mediators that are at a higher concentration than 02 and can therefore shuttle back and forth between the protein and the electrode faster than the enzyme is reduced, so that the arrival of the substrate such as glucose is always rate limiting. A typical chemical that works in this way is ferrocene, a sandwich of an iron cation between two cyclopentadienyl anions (Fig. 8). It exists in neutral and +1 oxidation states that are readily interconvertible at metal or carbon electrodes. A proprietary modified ferrocene is used in the aforementioned ExacTech

FIG. 8. The structure of the ferrocene-ferTOcinium ion couple that allows one to overcome the dependence of the glucose oxidase reaction on p02. The two iive-membered rings are cyclopentadienyl anions and the iron may he in either the Fe2+ or Fe'_ states, giving a total charge of 0 or + !.

carbon electrode-based glucose sensor. Other glucose oxidase-based electrodes have been employed on catheters for in vivo determination of blood glucose, with varying degrees of success (Gough et al., 1986). Thrombus formation is generally a problem, as is the possible alteration in localized glucose levels in tissue traumatized by insertion of probes, no matter how small. It may well be that use of the techniques employed in keeping pressure catheters clear will also work with biosensors such as this.

Optical Waveguide Sensors Fiber optics can be used either as thin flexible pipes to transport light to and from a sensor at a remote site, or in a way that takes advantage of the unique properties of optical waveguides. The former mode still dominates, and the CDI blood gas sensor uses three fibers just to move photons to and from the small volumes of immobilized chemistry at the probe end. The Schultz fiber optic glucose sensors involve a more sophisticated use of the light path exiting the optical fiber combined with clever use of lectin biochemistry. In principle, these sensors allow continuous measurement of blood glucose. There are at least two features specific to waveguides that have been used for sensors for in vitro measurement that may soon find themselves ready for in vivo use as well. In one, the ability of light sent down two fibers to interfere with itself on return to the source allows sensitive measurement of changes in the length or phase velocity of the fibers. This, in turn can be altered by enzymatically induced changes in the temperature of the fiber or its cladding in the volume surrounding the fiber. Another approach is to use the light in the evanscent wave that exists in the region just outside the waveguide to probe a small volume adjacent to the surface. If binding species such as antibodies are immobilized on the surface, it is possible to selectively excite and collect fluorescence from the surface layer even in the presence of high concentrations of fluorophore or other absorbers in the bulk solution. This technique has allowed the use of antibody-based detection of analytes such as theophylline in whole blood in a sensor designed by the ORD Corporation. These sensors are primarily for single use, and one fiber is used for each measurement. Nonspecific adsorption to the fiber surface, which is a serious interference in such sensors, can be reduced by using surface passivating films of proteins such as bovine serum album.

Acoustic/Mechanical Sensors The binding of a material to a surface changes its mass, which can change either the object's resonant frequency or the velocity of vibrations propagated through it. This has allowed development of sensors called surface acoustic wave (SAW) or bulk acoustic wave (BAW) detectors that are based on oscillating crystals. Sensitive detection of analytcs is relatively easy in rhe gas phase, and while there have been reports of selective detection of analytes using immobilized antibodies, there is still controversy as to how or if the technique works when the oscillating detector is in contact with liquid. It is, however, unlikely that this technique will prove applicable 10 in viva use, where some nonspecific adsorption of prorein is almost unavoidable.

Thermal and Phase Transition Sensors Chemical reactions can give up hear because they involve breaking and formation of cbcmical bonds, each of which has a characteristic enthalpy. There is also a strong effect of the heals of solution of the substrates and products, particularly charged species. Many eir/vmatic reactions release 25 to 10U kj/mol, or 5 to 25 kcal/ tnol (Table 6), A 1-mM solution of substrate completely eiizy-matically converted to product with a 5 kcal/mol heat of reaction would increase in temperature by 0.005°C, which is readily measureable in laboratory conditions. Sensors based on this principle arc in use as detectors in chromatography, and in principle could be applied to almost any enzymatic reaction. Some reactions have little or no heat production (e.g., ester hydrolysis, such as the acetylcholinesterase reaction) but can be observed using "tricks'' such as coupling the reaction to the heat or protonatioti of a buffer such as Tris:

Acetylcholine => H,CCO;H + choline ¡iH ~ 0 kj/mol H1CCO,H t- Tris ^H,CC07 * TrisH AH = -47 kj/mol.

Alternatively, a sequence of enzymes such as glucose oxidase followed by catalase can be used, which converts the hydrogen peroxide produced by the oxidase to O, and water in another exothermic reaction (Damelsson and Mosbach, 1987). However, the technical difficulties in making such measurements in rhe thermally noisy environment of the human body have so far prevented application of this principle to development of tn vivo sensors.

An alternative thermal approach is to use the depression in phase transition temperatures of pure compounds by dissolving

TABLE 6 Heat of Enzvmatic Reactions

Phase Transition Sensing volatil« in environment

TABLE 6 Heat of Enzvmatic Reactions



M7, (kl.'mol)



0 0

Post a comment