 |
Description  |
|
|
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates to peroxidase colloidal gold biosensors that provide
a detectable electrochemical response based on direct oxidation of a redox
protein at an electrode surface. In particular, mediatorless detection of
glucose is possible with colloidal gold adsorbed horseradish peroxidase in
the presence of glucose oxidase. The invention also includes methods of
mediatorless detection of various analytes and processes for the
preparation of colloidal gold adsorbed peroxidase based bioelectrodes.
2. Description of Related Art
Direct electron transfer between an enzyme and an electrode surface is of
practical as well as theoretical interest. An enzyme capable of direct
electron transfer immobilized on an electrode permits electrochemical
measurement of the enzyme substrate without addition of a mediator to the
analyte solution. Unfortunately, a serious problem with protein
electrochemistry is the slow mass transport process and strong adsorption
of protein molecules onto the electrode surface.
Because of the tendency of protein molecules to adsorb to surfaces, direct
electron transfer to or from the electrode surface is possible only for
the first layer of protein on the electrode. Even assuming a monolayer
coverage and completely reversible electrochemistry between the adsorbed
monolayer and the electrode surface, direct electron transfer between an
adsorbed monolayer of redox protein and an electrode surface would result
in a current approximately one-half that of the charging current.
While there are some examples of detectable electrochemical response based
on direct oxidation of a redox protein at an electrode surface, detection
has been difficult (Joensson and Gorton, 1989; Bowden et al., 1984).
Amplification of the signal can in some cases be achieved by adding enzyme
substrate.
Generally, in order to detect a signal, substrate is added in order to
induce enzyme turnover (Guo and Hill, 1991). This significantly amplifies
the signal which otherwise is generally too weak to be detected. A few
limited examples showing direct electron transfer between various enzymes
and electrode surfaces include cytochrome c peroxidase (Armstrong and
Lannon, 1987), p-cresolmethylhydroxylase (Gou and Hill, 1989), and
cytochrome c.sub.552 (Guo and Hill, 1990) at surface-modified electrodes
or in the presence of promoters. Other examples include cytochrome c
peroxidase irreversibly adsorbed on pyrolytic graphite (Paddock and
Bowden, 1989), and lysyl oxidase (Govindaraju et al., 1987) and
horseradish peroxidase (Joensson and Gorton, 1989) on spectrographic
graphite.
Current theories of non-mediated electrochemistry of proteins and enzymes
emphasize the importance of the electrode surface in facilitating direct
electron transfer (Guo and Hill, 1991). It has also been suggested that
direct electron transfer may proceed most easily to/from electrode
surfaces which provide an environment similar to the native environment of
the redox protein (Armstrong, 1991). However, there has been limited
success with approaches that deposit the redox protein directly on the
surface, presumably because of protein denaturation.
Horseradish peroxidase (HRP) has been suggested and studied as a
bioelectrode. An HRP electrode has high specific activity for H.sub.2
O.sub.2 with each H.sub.2 O molecule effectively converting ca. 25,000
H.sub.2 O.sub.2 molecules to H.sub.2 O per minute. In the presence of
H.sub.2 O.sub.2, HRP is efficiently converted to its oxidized form,
HRP.sub.ox (reaction (1)) (Frew et al., 1986). This can then be reduced,
as shown in reaction (2), either directly or through an electron transfer
mediator acting as an electron shuttle (Frew et al., 1986).
H.sub.2 O.sub.2 +HRP.sub.red .fwdarw.HRP.sub.ox +H.sub.2 O (1)
HRP.sub.ox +2e.sup.- .fwdarw.HRP.sub.red ( 2)
While electrodes based on horseradish peroxidase Will demonstrate direct
electron transfer (Joensson and Gorton, 1989), a major problem in
developing a redox system utilizing HRP has been to induce the
heterogeneous electron transfer step (reaction step 2) to proceed at a
reasonable rate. Acceptable rates of transfer are obtained in the presence
of a mediator, but without a mediator the rates are too slow to be of
practical value.
Biosensors are of particular interest for measuring glucose and there are
biosensors utilizing glucose oxidase as the sensing enzyme. A glucose
sensor based on gel immobilized glucose oxidase detects changes in pH when
coimmobilized with gluconolactase which hydrolyzes the lactone product of
glucose oxidation (Nakamoto, 1992). This type of glucose is, however,
relatively insensitive to glucose levels below about 0.1 mM.
More sensitive enzyme electrochemical sensor electrodes have been developed
that employ polymeric surface coatings. An enzyme such as glucose oxidase
dispersed in the polymer facilitates detection of hydrogen peroxide
produced during the reaction when employing a system incorporating a
reference/counter electrode with the enzyme-coated electrode (Rishpon et
al., 1992).
As a general principle, in the operation of a glucose biosensor, glucose
oxidase is reduced during the oxidation of glucose; the reduced enzyme is
then reoxidized either through an electron transfer mediator, which itself
becomes reoxidized on the electrode surface, or through molecular oxygen
present in the solution. The product resulting from oxygen reduction is
hydrogen peroxide which can be reoxidized at the electrode at high
positive potential, or, reduced to water at a high negative potential. In
either case, a high background signal is generated with high risk of
interferences from the sample matrix.
On the chemical level, a glucose biosensor is based on the conversion of
glucose (GO, the substrate or analyte) to gluconolactone (GL) in the
presence of a catalyst, glucose oxidase (GOD), represented by the
following equation:
GO+GOD.fwdarw.GL+GOD.sub.red ( 3)
In order to maintain continuous oxidation of GO, GOD.sub.red has to be
reoxidized to GOD. Equations 4-6 represent three different paths for
recycling GOD.
##EQU1##
The added electron transfer agent or mediator may be reoxidized as shown
in equation (7)
MED.sub.red -e.sup.- .fwdarw.MED.sub.ox ( 7)
Hydrogen peroxide generated from reduction of molecular oxygen will react,
depending on conditions, in the reduction mode, equation (8), or in the
oxidation mode, equation (9).
H.sub.2 O.sub.2 +2 e.sup.- +2 H.sup. .fwdarw.2 H.sub.2 O (8) reduction
mode
H.sub.2 O.sub.2 -2 e.sup.- .fwdarw.O.sub.2 +2 H.sup.+ ( 9) oxidation
mode
The process represented by equation (4) is normally very slow and therefore
considered impractical. The reaction with molecular oxygen, equation (5),
will take place unless oxygen is purged from the system. Mediated
reactions, represented by equation (6), can be quite efficient, depending
on the mediator.
For purposes of developing a practical glucose biosensor, three options
would include, based on equations 3-9:
Mode one: (3).fwdarw.(6).fwdarw.(7): oxidation mode;
Mode two: (3).fwdarw.(5).fwdarw.(9): oxidation mode; and
Mode three: (3).fwdarw.(5).fwdarw.(8): reduction mode
Mode one operates at a potential of 0.3-0.4 V and has the advantage of
being a direct measure of the glucose oxidase redox process. There are,
however, several disadvantages, including requirement of a mediator which
to be effective must be immobilized near the electrode surface. The
effectiveness, operational potential (0.3-0.4V/Ag/AgCl) and the background
current depend on the mediator. Moreover, the mediator must be initially
in its oxidized form in order to minimize the initial background current.
Unfortunately, good mediators, e.g., ferrocene and its derivatives, are
only readily available in their reduced form.
Yet another disadvantage of Mode one operation is sensitivity to molecular
oxygen. O.sub.2, when present, will compete with the mediator. As a
practical matter, purging the oxygen is time-consuming and expensive in
large scale operations. The effect of O.sub.2 depends on the relative rate
of the reactions shown in equations (5) and (6). A further disadvantage is
the dependence of the O.sub.2 effect on glucose concentration as well as
the concentration of molecular oxygen present. Variation of ambient
O.sub.2 concentration therefore will have unpredictable effects on the
mediated signal. Even at constant O.sub.2 concentration, predictability is
difficult because the effect is more detrimental at low glucose
concentrations than at higher glucose concentrations (Hale et al., 1991;
Gregg and Heller, 1990). At present, no mediators have been reported that
operate efficiently enough to eliminate the oxygen effect.
Mode two operates at a potential of 0.6-0.7V and has several advantages,
including the fact it is not sensitive to oxygen at low glucose
concentrations as there is usually sufficient oxygen in the solution.
Additionally, a mediator is not required and there are no competitive
reactions, assuming no interfering substances are added in the sample.
Mode two does, however, have several disadvantages. The process is
sensitive to oxygen at high glucose concentrations when oxygen which is
normally present may become limited. The product, not the enzyme redox
process, is measured. And the high operational potential, 0.6-0.7
V/Ag/AgCl, results in a high background current, so that the signal
current may be difficult to detect.
Mode three operates at 0V Ag/AgCl and has a number of advantages. This
system can be coupled to HRP with direct electron transfer in the
reduction mode, equation (8), at 0V on the electrode. As in Mode two, no
mediator is required, there are no competing reactions and there is no
oxygen sensitivity at low glucose concentrations. A distinct advantage is
low background and interference due to the low operational potential.
Mode three disadvantages include sensitivity to oxygen at high glucose
concentrations and measurement of a product rather than the enzyme redox
process directly. Additionally, two enzymes are required, adding
complexity to the system and possible additional expense for fabrication.
Enzyme electrochemical sensors for glucose determination have been
described (Rishpon et al., 1992). In these Mode one type biosensors, GOD
is incorporated into membranes near the electrode surface to reduce
interference from undesired oxidizable compounds and to reduce oxygen
sensitivity. The electrode is however not sensitive to glucose
concentrations below about 1 mM.
Electron transfer agents, such as ferrocenes, have been used in conjunction
with glucose oxidase. However, two major drawbacks exist. In common
practice, electron transfer mediators are small molecules, typically
ferrocene for glucose oxidase based biosensors. It is generally desirable
to immobilize a mediator to keep it close to the surface; however, small
molecules are difficult to immobilize. A more difficult problem is the
ubiquitous presence of molecular oxygen. Oxygen will always be reduced to
some extent, even in the presence of a mediator. The result is that, while
a mediated response may produce a satisfactory response to relatively high
glucose concentrations, it is not feasible to measure low glucose (100
.mu.M range) concentrations because of background current and the effect
of oxygen.
SUMMARY OF THE INVENTION
The present invention addresses one or more of the foregoing problems in
providing novel biosensors operating on direct electron transfer arising
from reduction of a colloidal gold immobilized peroxidase deposited on a
conducting surface. The bioelectrodes of the present invention when
suitably coupled with a transducer are capable of detecting a current
generated from reaction of hydrogen peroxide with the peroxidase on the
conducting surface of the biosensor. Hydrogen peroxide, produced in the
presence of oxygen during an oxidase catalyzed reaction of an appropriate
substrate, efficiently oxidizes surface-deposited colloidal gold adsorbed
peroxidases. The disclosed bioelectrodes prepared with colloidal gold
adsorbed horseradish peroxidase and glucose oxidase are particularly
suitable for determination of glucose. Current is produced in the presence
of glucose at glucose concentrations as low as 1 .mu.M. Other oxidases,
such as alcohol oxidase, galactose oxidase, lactic acid oxidase, amino
acid oxidase, cholesterol oxidase, xanthine oxidase and the like are also
useful in practicing the invention so long as hydrogen peroxide is
generated during the catalytic reaction.
The invention relates to novel colloidal gold based bioelectrodes employing
an immobilized colloidal gold adsorbed peroxidase and at least one other
enzyme in the oxidase class. In preferred practice, a bioelectrode is
prepared from horseradish peroxidase which is first adsorbed onto
colloidal gold sol particles and then deposited on a conducting electrode
surface. One or more oxidase enzymes are then added prior to determination
of selected analytes. The oxidase, selective for a particular analyte, may
be added to a sample solution in soluble or immobilized form or,
preferably, immobilized near or on the conducting surface where the
peroxidase is deposited. Whether localized on or near the conducting
surface, it is understood that the oxidase will be capable of coupling
with HRP and as such to be "in communication" with a conducting surface so
as to operate in the reduction mode (Mode three) previously herein
described in equations (3), (5) and (8).
Horseradish peroxidase has a very specific activity toward hydrogen
peroxide and is efficiently converted to its oxidized form. As disclosed
herein, horseradish peroxide present on a conducting electrode surface can
be efficiently reduced directly on the electrode surface at a voltage near
0 volts Ag/AgCl. This takes place through a direct electron transfer and
does not require an electron transfer mediator. A novel aspect of the
invention is the absorption of horseradish peroxidase on to the surface of
colloidal gold particles prior to depositing the enzyme on a conducting
electrode surface.
While the invention has been demonstrated with horseradish peroxidase, it
is understood that other sources of peroxidases may also be employed, not
necessarily limited to horseradish. Moreover, other peroxidase-type
enzymes are contemplated as useful so long as the enzyme will accept
hydrogen peroxide as a substrate. Depending on the particular bioelectrode
desired, there are several properties to be considered. Properties such as
enzyme stability, high specific activity, and efficient conversion of
hydrogen peroxide are factors to consider. The invention need not be
limited to immobilized native peroxidases. Genetically engineered,
truncated enzymes including active catalytic sites, or modified
catalytically active species may also be useful and even more efficient in
some applications.
A second component of the disclosed bioelectrode includes an oxidase. As
used herein, oxidase includes any enzyme that is capable of generating
hydrogen peroxide during a catalytic reaction. The oxidase will be
selected to catalyze a reaction with a desired analyte. By analyte is
meant a substrate for the selected enzyme. A further constraint in forming
an operable system is the presence of molecular oxygen which, during the
catalytic reaction, will be converted to hydrogen peroxide which will
oxidize colloidal gold adsorbed horseradish peroxidase located on the
electrode surface. When appropriately combined with a reference/counter
electrode, direct electron transfer occurs at the electrode surface
resulting in regeneration of the reduced form of horseradish peroxidase.
Bioelectrodes of the invention are basically two-enzyme electrodes. A
sensing enzyme, typically horseradish peroxidase, is adsorbed to the
surface of colloidal gold sol particles. Adsorption to the surface of
colloidal gold particles appears to stabilize the enzyme and to provide a
conducting matrix. In practice, colloidal gold adsorbed HRP is deposited
on a conducting electrode surface. Deposition may be through spraying,
dipping, electrodeposition, solvent evaporation or a variety of other
well-known techniques but is most conveniently accomplished by simply
evaporating a colloidal gold adsorbed horseradish peroxidase solution onto
the electrode surface. An oxidase is provided to detect a desired analyte.
Examples of analytes include cholesterol, xanthine, monosaccharides such
as glucose, amino acids and alcohols. The oxidase selected will, however,
produce hydrogen peroxide during catalytic conversion of a desired
analyte. The hydrogen peroxide produced is detectable by the peroxidase
immobilized on the conducting electrode surface.
Enzymes employed in conjunction with horseradish peroxidase typically
include oxidases. Such enzymes generate hydrogen peroxide from molecular
oxygen in the course of the catalytic reaction. Preferred enzymes include
cholesterol oxidase, amino acid oxidase, alcohol oxidase, lactic acid
oxidase, galactose oxidase and, most preferably, glucose oxidase.
In yet another aspect of the invention colloidal gold adsorbed enzyme is
first immobilized in a matrix which is then positioned at or near the
electrode surface. Several types of matrix are suitable, including
hydrophilic polymers such as the carrageenans, agar and similar
hydrophilic gels. The selected matrix may be used merely to protect the
surface of the gel, or alternatively as a second immobilization material
in which, for example, colloidal gold adsorbed enzyme is dispersed. It is
contemplated that more than one enzyme may be conveniently dispersed
within a gel matrix. Appropriate alteration of the electrode potential
when configured as a biosensor allows sequential determination of more
than one analyte.
Another aspect of the present invention includes a method for enzyme
electrochemical detection of a desired analyte. A bioelectrode as
described herein is contacted with a sample that may contain the analyte
of interest. Analyte present in the sample is determined from the amount
of current generated from hydrogen peroxide reduction by peroxidase
immobilized at the conducting surface. Hydrogen peroxide produced during
the enzyme catalyzed analyte conversion is selectively reduced to water by
horseradish peroxidase on the electrode surface. All electron transfers in
the disclosed systems will operate without addition of electron transfer
mediators, and will do so more efficiently than when mediators are present
when the appropriate methods of preparation are employed. However, this
does not preclude the use of a mediator if desired. In some
configurations, the use of a mediator may offer more efficient transfer.
Analytes to be analyzed by the present invention may be found in a wide
variety of aqueous samples including water, urine, blood, sweat, and other
body fluids such as vaginal or seminal fluids. In a most preferred
embodiment, a horseradish peroxidase/glucose oxidase bioelectrode will
detect glucose by direct electron transfer.
Yet another aspect of the invention is a selective bioelectrode for the
detection of glucose. Such a bioelectrode includes a first layer of
colloidal gold-adsorbed horseradish peroxidase deposited on a conducting
electrode surface and a second layer of colloidal gold-adsorbed glucose
oxidase preferably overlying the first layer of colloidal gold-adsorbed
HRP. A preferred conducting electrode surface is glassy carbon, although
usable conducting surfaces include carbon, gold, platinum, and the like.
In preferred practice both layers of colloidal gold adsorbed enzymes are
evaporatively deposited onto a conducting surface. Typically, glucose
oxidase immobilized on colloidal gold will be in contact with the
colloidal gold adsorbed HRP. Glucose bioelectrodes constructed in this
manner are capable of detecting glucose levels at least as low as 1 .mu.M
and generally show a linear response to glucose concentrations as high as
250 .mu.M.
Bioelectrodes convenient for detecting glucose are typically constructed as
biosensors by combining with reference and counter electrodes. Samples
containing glucose or suspected of containing glucose are contacted with
the bioelectrode and the amount of current produced is related to the
amount of glucose present. Current is produced by reduction of hydrogen
peroxide at the conducting surface of the bioelectrode and is typically
measured at 0 volts/Ag versus Ag/AgCl.
A novel aspect of the present invention is the capacity of the disclosed
bioelectrodes to amperometrically detect a selected analyte by direct
electron transfer at the electrode surface without the need for an
electron transfer mediator. Surface contact of the detecting enzyme,
typically horseradish peroxidase, and, surprisingly, the coating
distribution of the enzyme on the surface of the colloidal gold particles
contribute to the reactivity and response of the electrode. In general,
monolayer coverage of the colloidal gold particle surface by HRP appears
to provide the most effective electron transfer without a mediator. This
does not preclude effective mediatorless electron transfer with less than
monolayer coverage or even imperfect or partial coatings. This likely
depends on the enzyme adsorbed to the colloidal gold as well as the
properties, e.g. size, of the sol particles to some extent. Additional
layers of HRP on the surfaces of colloidal gold particles, at least where
a glucose bioelectrode is concerned, do not increase response. Surface
coverage significantly greater than monolayer may generally inhibit direct
electron transfer response.
BRlEF DESCRIPTION OF THE DRAWINGS
FIG. 1 shows the steady-state amperometry measurement of current vs.
peroxide concentration in stirred buffer with or without 0.2 mM
ferrocenecarboxylic acid mediator at 0V/Ag. ) HRP solution evaporated
onto glassy carbon without mediator; ) HRP solution evaporated onto
glassy carbon with a mediator in solution; ) Au-HRP sol evaporated onto
glassy carbon without a mediator.
FIG. 2 is the same measurement as in FIG. 1 except that colloidal gold
immobilized HRP was deposited on a gold film over a glass electrode
surface rather than on glassy carbon.
FIG. 3 is a plot of the activity ratio (electrode response in presence of
azide/electrode response in absence of azide) vs. sodium azide
concentration.
FIG. 4 shows a plot of steady state current in nA versus micromolar
concentration of hydrogen peroxide for a colloidal gold/HRP electrode.
Response in acetate, (.DELTA.) and phosphate ( ) buffers in the absence of
an electron transfer mediator is shown.
FIG. 5 shows current generated relative to glucose concentration. The
detecting bioelectrode was constructed from a colloidal gold/HRP deposited
layer underneath a colloidal gold glucose oxidase layer. Measurements were
made in a microcell without a mediator. The effect of pH (acetate
(.DELTA.) and phosphate ( ) buffers) is indicated.
FIG. 6 is an amplified portion of the plot shown in FIG. 5 in the region up
to 300 .mu.M glucose concentration.
DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS
The electrochemistry associated with redox enzymes and low molecular weight
proteins attached to electrode surfaces has become an important and
well-established area of research (Cass, 1986 . Most reports in the
literature describe the electrochemistry of small redox proteins such as
cytochrome c (Albery et al., 198I) and ferredoxin (Hagen, 1989), generally
on surface-modified electrodes, or mediated electron transfer between an
enzyme and an electrode surface (e.g., glucose oxidase on a graphite
electrode with a ferrocene derivative used as mediator (Cass et al.,
1990)).
Colloidal gold has been used as an immobilization support for proteins
employed as markers for electron microscopy. Under these conditions, the
biological macromolecules are known to retain biological activity. It has
been shown that several enzymes maintain their enzymatic and
electrochemical activity when immobilized on colloidal gold (Henkens et
al., 1987). The immobilization of a redox enzyme on colloidal gold may
help the protein assume a favored orientation and/or provide conducting
channels between prosthetic groups and the electrode surface. In either
situation, effective electron transfer distance is reduced, thereby
facilitating charge transfer.
Although it is contemplated that any of a number of peroxidases may be
employed as an electron transfer enzyme at the conducting surface of the
bioelectrode, horseradish peroxidase is particularly preferred and has
been used to demonstrate the invention. Horseradish peroxidase (HRP) is
available in pure form with very high specific activity, and has been
successfully immobilized on colloidal gold with retention of its
biological activity (Henkens et al., 1991). The mediated reduction of
immobilized HRP occurs at 0V (Ag/AgCl) at a reasonable rate.
The reduction of native HRP to its ferrous form at modified electrodes is
effected at relatively high negative potentials (-0.71V/SCE). The
reduction of oxidized HRP (eq. 2) to its native state will proceed in the
absence of a mediator at a glassy carbon electrode as well as in the
presence of the mediator ferrocenecarboxylic acid (Frew et al. 1986). This
demonstrates the thermodynamic possibility of direct reduction of HRP on
the electrode surface near 0 V/Ag. In the absence of an electron transfer
mediator, current measured by cyclic voltammetry or steady-state
amperometry using an HRP modified electrode in the presence or absence of
H.sub.2 O.sub.2 is an indication of direct HRP reduction.
High background current problems associated with mediated electron transfer
of glucose oxidase at the electrode are overcome with the use of
horseradish peroxidase as part of a dual enzyme electrode for sensing low
concentrations of analytes. Horseradish peroxidase may be co-immobilized
for peroxide reduction at ideal potentials (0V vs Ag/AgCl). Background and
interferences from the sample matrix are eliminated or greatly reduced. A
dual enzyme HRP/glucose oxidase electrode will be less sensitive to oxygen
and variations in oxygen concentration at low glucose concentrations with
less effect on the signal current than at higher concentrations of
glucose.
Materials and Methods
HRP, type VIA, was purchased from Sigma Chemical Co. (St. Louis, Mo.) and
dialyzed before use against 2 mM sodium phosphate at pH 7.0 or used
directly without further treatment.
Gold trichloride (HAuCl.sub.4.3H.sub.2 O) was purchased from Fisher
Chemical Co.; ferrocenecarboxylic acid was from Aldrich.
Preparation of Gold Sols
Gold sols were prepared with a particle diameter of approximately 30 nm. A
solution of 0.3% aqueous sodium citrate was added to a boiling rapidly
stirred solution of 0.01% gold trichloride and the solution refluxed for
30 min. The final concentrations (w/v) were 0.01% HAuCl.sub.4 and 0.03%
sodium citrate. The particle size was estimated by filtration of the sol
through polycarbonate membranes (Nuclepore Corp.) of varying pore sizes
using an Amicon micro ultrafiltration unit. Approximately 40% of the sol
passed through a 500 .ANG. Nuclepore filter and was quantitatively
collected on a 300 .ANG. Nuclepore filter.
Preparation of Colloidal Gold Adsorbed Enzyme
The gold sol was concentrated by centrifugation at room temperature. The
concentrated sol was mixed with appropriate amounts of HRP and a fixed
amount of the Au-HRP sol evaporated on a coplanar carbon electrode
surface. HRP concentration in the sol measured against electrode activity
was used to determine optimum composition of the Au-HRP sol.
Electrodes
Glassy carbon electrodes were prepared by inserting a glassy carbon rod
into a hot, soft teflon cylinder with a copper or stainless steel rod
connection. On cooling, the teflon became tightly wrapped around the
glassy carbon rod. Silver epoxy served to connect the metal with the
glassy carbon.
Coplanar glassy carbon electrodes were prepared by first wrapping a glassy
carbon rod with heat shrinkable tubing. A silver wire served as a
reference electrode and a platinum foil as a counter electrode, each
wrapped with additional heat shrinkable tubing with at least one layer of
insulating tubing between each of the three electrodes. All three
electrodes were exposed on the same surface.
Two electrode materials were used to prepare four different HRP or HRP
colloidal gold (HRP-Au) modified electrodes. One consisted of three vapor
deposited gold strips on glass (Au/glass). Silver was electroplated onto
one of the strips as the reference electrode. HRP solution or HRP-Au sol
(3 .mu.l) was evaporated onto one of the strips to make the HRP-Au or HRP
modified working electrode. The remaining bare strip served as the
auxiliary electrode. The second HRP electrode configuration consisted of a
glassy carbon working electrode, Ag/AgCl reference and Pt wire auxiliary
electrode. HRP-Au sol or HRP solution was evaporated onto the working
electrode. Three-electrode cells were used with 5 mL sample size. The HRP
coated electrode surfaces had an area of about 7 mm.sup.2.
The buffer solution was 50 mM phosphate at pH 6.8 with 10 mM KCl unless
otherwise specified. No deaeration was necessary in most cases. In
mediated experiments, ferrocenecarboxylic acid was used at a concentration
of 0.20 mM, which is well into the region where the electrode response is
independent of the mediator concentration.
A Pine Instrument RD4 bi-potentiostat interfaced to an IBM-386 computer was
used for the measurements. The system was controlled with ASYST programs,
and electrochemical data were directly collected and processed in the
computer. Cyclic voltammograms were obtained without stirring the
solution. In steady-state amperometry experiments the potential was set at
0V/Ag in stirred buffer, and the steady state current was measured.
Although there are examples in the literature of direct electron transfer
between a redox protein and an electrode, facile transfer of electrons has
generally been considered difficult with a non-functionalized electrode
surface. HRP adsorbed on flat Au/glass did not respond to H.sub.2 O.sub.2
unless an electron transfer mediator was present, indicating the lack of
direct reduction of HRP. However, when HRP was adsorbed to a colloidal
gold sol and then deposited on glassy carbon (FIG. 1 ()) or a flat metal
surface (FIG. 2 ()), direct reduction of HRP on the electrode surface was
observed.
A freshly polished glassy carbon surface has many functional groups, so
that in a sense a chemically modified surface is exposed to the solution
(Kinoshita, 1989). The functional groups may act as absorption sites which
promote electron transfer. HRP adsorbed onto freshly polished glassy
carbon catalyzed reduction of H.sub.2 O.sub.2 without a mediator (FIG. 1,
). Gold film surfaces, which do not have functional groups such as are
found on the surface of glassy carbon, did not catalyze the direct
reduction of H.sub.2 O.sub.2 (FIG. 2, ). Reduction proceeded reasonably
well in the presence of a me (FIG. 2, ).
Electrodes prepared by deposition of a HRP-Au sol onto glassy carbon (FIG.
1 () or onto Au on glass (FIG. 2 ()) responded to H.sub.2 O.sub.2 in a
stir-rate dependent manner at low peroxide concentrations (<50 .mu.M
H.sub.2 O.sub.2) In this region of the curve the reaction was mainly
diffusion controlled so that diffusion of the substrate to the electrode
surface was the rate-limiting step.
At high peroxide concentrations (>150 .mu.M H.sub.2 O.sub.2) the response
was enzymatically controlled, indicated by the lack of dependence on
stirring rate. In this region of the curve, addition of mediator to the
analyte solution increased the response significantly, i.e., addition of a
mediator extended the linear range. This may arise because some of the
colloidal gold adsorbed HRP molecules are not in an appropriate
orientation for direct electron transfer. Addition of a mediator allows
more of the adsorbed HRP molecules to participate in the electron transfer
reaction, increasing the response.
Under optimum conditions when HRP coatings on the colloidal gold particle
surfaces are equivalent to a monolayer coverage or less, there are no
apparent mediator effects and the direct electron transfer if proportional
to HRP loading.
For simple absorption of HRP on a glassy carbon electrode it may be assumed
that only the first layer of the adsorbed HRP molecules accepts electrons
directly from the electrode surface. Direct electron transfer from the
electrode surface to the second layer and beyond may be neglected due to
the long electron transfer distances involved and the specific
orientations required for electron hopping or self exchange to occur.
However, the presence of an electron transfer mediator should promote
efficient charge transfer beyond the first layer. The observation that
addition of mediator gives only a small increase in catalytic current in
the case of simple absorption on a flat glassy carbon surface (FIG. 1,
.quadrature.) indicated that no more than a monolayer of HRP was adsorbed
on the surface. Determination of the amount of adsorbed protein based on
the total enzymatic activity of HRP on the electrode indicated that the
surface coverage was<5% of a monolayer. The small increase in signal upon
addition of a mediator also implied that most of the adsorbed protein
molecules had good access to electrode surface functional groups. The
absorption appeared to be specific and uniform.
A colloidal gold surface is very different from flat bulk gold. Although
the exact nature of the colloidal gold/protein/electrode surface
interaction has not been completely defined, there are several ways in
which colloidal gold may be visualized as assisting in electron transfer
between a redox protein and an electrode surface. Colloidal gold particles
have high surface to volume ratios. Uncontaminated gold sol particle
surfaces have high surface energy and so are very active. The interaction
with protein molecules can be very strong. The small size of the colloidal
gold particles (approximately 30 nm) gives the protein molecules more
freedom in orientation thus increasing the possibility that the prosthetic
group is closer to the metal particle surface. The distance between the
protein and the metal particles is shorter, facilitating charge transfer.
When colloidal gold adsorbed HRP is deposited onto an electrode surface,
HRP coated colloidal gold particles function as electron-conducting
pathways between the prosthetic groups and the electrode surface.
The larger effective surface area of a colloidal gold particle may allow
more enzyme molecules to be immobilized at or near the electrode surface.
The possibility for multilayers of effective Au-HRP layers may be another
mechanism by which the signal from colloidal gold assisted immobilization
is increased. However, the effective layer should not be too deep because
the signal does not increase proportionally with the amount of HRP-Au sol
deposited (1-10 .mu.l on a 3 mm diameter glassy carbon surface), with or
without an electron transfer mediator. Unmediated electron transfer
decreases when the deposited HRP-Au layer is too thick, probably because
interior enzyme-Au layers are less efficient conductors than glassy
carbon.
Assuming that the average diameter of the sol particle is 30 nm and the
density is 17.0 g/ml, then 3 .mu.l of 7.5 mg Au/ml sol deposited onto a
glassy carbon surface of 3 mm in diameter is equivalent to about 12 layers
of Au sol particles. This surface coverage gives the best performance,
both with or without a mediator. Additional Au layers cause some
deterioration in unmediated response but there is little effect on
mediated response. This suggests that the deposited Au layers are not very
porous and that the accessible depth is about 12 layers of deposited Au
sol. Even within the 12 layers, only the outermost layers are important,
because changing from 4 to 1 2 layers increased the signal by only 10-20%
with or without a mediator. In consideration of both the electrode
performance and cost, 3 .mu.l HRP-Au sol is optimum for a 3 mm diameter
glassy carbon surface.
Although only the outermost gold layers appear to contribute the major
portion of the accessible enzyme molecules, enzyme loading and mediator
effects with colloidal gold assisted immobilization are significantly
higher when compared with simple absorption on surfaces. Spectroscopic
data for the enzymatic activity of HRP adsorbed on colloidal gold before
deposition on the electrode surface indicate that the active enzyme
coverage on the gold sol particle surfaces is about 40% of a theoretical
compact monolayer. This is consistent with absorption of .gamma.-globulin
onto latex particles (Fair and Jamieson, 1980). Multilayer absorption of
protein molecules on a solid support surface is likely negligible. If
absorption is not specific, protein molecules may have multiple
orientations on the surface. The strong interactions between the protein
and the Au sol surface may increase the surface density of the adsorbed
protein, and some of the restricted orientations may also favor direct
electron transfer between protein molecules and the conductor surface. It
is likely that all of the active enzyme molecules are on the first layer
of the adsorbed surface, but only part of the molecules have the correct
orientation for direct electron transfer.
EXAMPLE 1
This example illustrates that colloidal gold adsorbed HRP deposited onto a
glassy carbon surface produces an excellent electrode response to hydrogen
peroxide without a mediator. Colloidal gold or HRP alone elicited either
no response or a very low response.
Colloidal Gold Deposited on Electrode Surface
Colloidal gold sols were prepared and evaporated onto glassy carbon or
Au/glass surfaces. None of the electrodes tested had a significant
response to H.sub.2 O.sub.2 (up to 200 .mu.M) in steady-state amperometry
measurements at 0V/Ag. The sensitivity was<0.05 nA/.mu.M H.sub.2 O.sub.2
in the presence or absence of ferrocenecarboxylic acid. Electrodes
prepared from colloidal gold deposited on glassy carbon showed no
catalytic current in cyclic voltammograms recorded with 0 to 2 mM H.sub.2
O.sub.2.
HRP Deposited on Electrode Surface
Aliquots of a solution of HRP in buffer with no colloidal gold were
evaporated onto the surface of Au-glass or glassy carbon electrodes and
the response of these electrodes to H.sub.2 O.sub.2 measured by
steady-state amperometry. The HRP/glassy carbon electrode showed a low
response to H.sub.2 O.sub.2 in the absence of an electron transfer
mediator (FIG. 1, ). Addition of ferrocenecarboxylic acid slightly
increased the response to H.sub.2 O.sub.2 (FIG. 1, ). The HRP/Au/glass
electrode gave very little response to H.sub.2 O.sub.2 in the absence of
ferrocenecarboxylic acid (FIG. 2, ), but gave an improved response in the
presence of the mediator (FIG. 2, .quadrature.).
Colloidal Gold-HRP Deposited on Electrode Surface
Electrodes were prepare | | |
