dIN.
CHEM.
26/11,
1569-1572 (1980)
Antigen-AntibodyReactionInvestigatedwith Use of a ChemicallyModified
Electrode
Naoto Yamamoto,1 Yoshikatsu Nagasawa,’ Sadanobu Shuto,1 Hiroshi Tsubomura,’ Masanobu Sawal,2
and Hajime Okumura2
The antigen-antibody reaction of human choriogonadotropin has been potentiometricaliy investigated by use of
a cyanogen bromide-treated electrode coated with the
corresponding anti-serum. The potential of the modified
electrode shifts in the positive direction upon contact with
a solution of choriogonadotropin. The rate of the reaction
at the interface between the electrode and the solution is
estimated from the potentiometric measurement to be of
the order of 14 mor1 s1 in diethyl barbiturate buffer,
depending to some extent on the rotation speed of the
stirrer. The change in potential is almost proportional to
choriogonadotropin concentration. It is also pH dependent,
the maximum response being at pH 8.7. The technique,
applied to samples of human urine, has shown a specific
response to chorlogonadotropin.
Additional
urine
choriogonadotropin
pregnancy detection
Keyphrases
potent lomefry
#{149}
The highly specific nature
of complex-forming
reactions
between biologicalsubstances such as antigen and antibody
is well recognized. Recently, some new methods have been
proposed for detecting these substances by virtue of such reactions (1-3). A potentiometric method for detecting biological substances has also been developed, with which the reaction products are monitored by measuring the potential of
a glass electrode that is covered with an enzyme-immobilized
membrane (4-11). We have also investigated antigen-antibody and enzyme-inhibitor reactions by measuring the change
of potential of a chemically modified electrode resulting from
reactions at the electrode-solution
interface (12, 13). The
changes in potential have been interpreted by use of a model
of an electric double layer on the electrode surface. Indeed,
the electrical changes caused by the reactions between biological substances might be closely related to the fundamental
mechanisms of the physiological actions in life.
Extensive experimental studies of surface potentials of
solid-gas and liquid-gas interfaces have been made in connection with catalytic reactions and adsorption phenomena.
Clarification of electrical properties at the solid-solution
interface is very important
for comprehensive
understanding
of electrochemical
reactions, but few such potentiometric
investigations have been made so far.
Here we report our experimental
results on the reaction rate
and the pH dependence of an antigen-antibody
reaction occurring at the surface of a modified titanium electrode.
Materials and Methods
The methods for preparing human choriogonadotropin
(hCG) and anti-serum
to it (anti-hCG) in rabbits are described
elsewhere (12-14). The relative molecular masses were found
‘Department of Chemistry, Faculty ofEngineering Science, Osaka
University,
Toyonaka, Osaka, 560 Japan.
2ResearchLaboratories,
Teikoku Hormone Mfg. Co., Ltd., Kawasaki-shi 213, Japan.
Received Nov. 26, 1979; accepted July 1,1980.
to be 40 680 for hCG (15) and 160 000 for anti-hCG (16). The
electrodes were prepared in a way similar to that described
elsewhere (12, 13). Briefly, the procedure is as follows. The tip
of a titanium wire, 1 mm in diameter, is heated at about 1000
#{176}C,
to form a thin oxide layer on its surface. The wire is then
inserted into a glass tube and cemented with epoxy resin as
shown in Figure la, The protruding
tip, 1.5cm long, is washed
with distilled water, and chemically activated in a stirred
aqueous solution of cyanogen bromide for 20 mm, with the pH
maintained between 10 and 11 by dropwise addition of a 1.0
mol/L sodium hydroxide solution. The tip is then washed with
distilled water and immersed for 40 mm in aO.1 mol/L sodium
bicarbonate solution (pH 8.4) containing anti-hCG,
then
soaked in a 1.0 mol/L urea solution for 20 mm to deactivate
the remaining active sites, in order to prevent excessive adsorption of protein. A urea electrode, used as the reference
electrode, is prepared by a similar chemical treatment of the
electrode with cyanogen bromide and urea.
Differences in electrical potential between the anti-hCG
electrode and the urea electrode were usually measured in 50
mmolfL diethylbarbiturate
buffer solutions (6 mL, pH 8.6)
with use of a vibrating-reed
electrometer
(Model TR 84M;
Takeda Riken Industry Co., Tokyo 176, Japan) at 35 #{176}C
(Figure lb). The buffer solution was stirred with a Teflon
propeller, rotated
by a pulse motor at an appropriate
speed.
Results and Discussion
Figure 2 shows typical results for an anti-hCG electrode
potential vs the reference urea electrode. As was described in
previous papers (12,13), the potential begins to shift toward
the more positive when a small amount of hCG solution is
added to the buffer solution. When the hCG solution is replaced with a buffer containing
no hCG, the electrode resumes
the same potential as initially. It was clarified in the previous
paper that the change of potential is the result of the specific
reaction between the anti-hCG, which is fixed on the electrode
surface, and the hCG in the solution. It has been found in the
present work that the reaction product-i.e.,
the antigenantibody
complex-dissociates
into anti-hCG and hCG when
the electrode is dipped into a cold solution of hydrogen chloride, pH 1.5, for 2 mm, and the potential returns to its initial
value. The electrode will then show a similar response in potential on exposure to hCG again, and the hCG reaction and
detachment with HC1 can be repeated several times, the recovery of the electrode response being about 85% after each
such repetition.
As discussed previously (12, 13), the antigen-antibody
reaction can be expressed by the following formula.
k
A + B
-
P
(1)
where A is the antibody chemically bound on the electrode
surface, B is the antigen present in the solution, and P is the
reaction product on the electrode surface. The rate constant,
k, is naturally considered to depend on the thickness of an
unstirred layer on the surface (17). The amount of anti-hCG
CLINICALCHEMISTRY,
Vol. 26, No. 11, 1980 1569
-10
titanium
>
E
glass
-12
spoxycement
-f-
1.5 c
-14
antl-hCG
modifIed
(a)
0
(b)
Fig. 1. Diagram of the antl-hCG modified electrode (a) and the
measuring cell (b)
attached on the working electrodeismuch smaller than the
amount of hCG contained in the solution under our experimental conditions. Therefore, by assuming that the potential
shift is proportional
to the quantity of the reaction product
on the electrode surface, the potential at time t, U(t), is expressed as a function of time, t, after addition of the antigen
as follows:
U(t)
=
(U1
-
Uf) exp (-at) + U1
dt
=
10-D1’
dx
=
103D
EBb] [B1]
-
#{163}
10-3D EBb) (3)
#{163}
where n is the amount of reactant B transported in a cross
section of 1 cm2, D the diffusion constant of B, x the distance
from the interface, #{128}
the thickness of the unstirred layer, and
EBb]and [B1) are the concentrations of B (in units of mol L)
in the bulk solution and at the interface, respectively. The
diffusion constant of hCG reportedly is 8 X 10 cm2 s’ at 20
0C (18). Assuming, as usual, that the value of #{163}
is i0
cm in
the region of higher rotating speeds (17), the value of dn/dt
is estimated to be about 8 X 10 EBb] (L s’ cm-2).
The surface density of the reaction product formed at the
interface, n mol cm2, is given by the following equation,
based on equation 1:
=k(nAo-np)[B]
FIg. 2. The behavior of the
barbiturate buffer solution
(4)
1570
CLINICAL CHEMISTRY,
Vol. 26, No. 11, 1980
30
40
50
Time(mini
electrode in a 50 mmol/L
antl-hCG
of Figure 3 for the well-stirred solutions, the amount, nB, of
the reactant B consumed by the reaction is given as follows:
x 10-7[B]
1.6
=
dt
dt
Thus, the rate of consumption is much smaller than would be
expected for a diffusion-controlled process. Consequently, the
presence of the unstirred layer does not affect the reaction rate
for the case of the well-stirred solutions (Figure 3).
Figure 4 shows the effect of the concentration
of added hCG
on the response profile of the electrode. The experiment was
performed with the stirrer rotating at 300 rpm. it is found from
Figure 4 that the slope of the response curve increases with
the concentration of hCG added. In fact, the response profile
is essentially described by first-order kinetics.
Figure 5 shows the relationship
between the a value derived
from equation 2 and the concentration of hCG in the reaction
cell. The a value is approximately
proportional
to hCG c#{244}ncentration, suggesting that the concentration
of the sample
can be determined by use of the anti-hCG electrode. The rate
constant, k, on the average is derived from the slope to be 1.2
X 10’ mol’ s,
nearly equal to the above-mentioned
value
1.6 x 10 mo!1 s’ and that calculated
from the previous
papers (12, 13) 1.25 X 10’ mol1
As far as we are aware, no reaction rate constant for hCG
1.5
U,
1.0
-I
x
0.5
0
0
where nAo is the surface density of the reactant A attached on
the electrode surface, and k is the reaction rate constant. The
surface density of anti-hCG at the electrode surface was estimated to be about 10-11 mol cm2 in the previous papers(12,
13). Taking k to be 1.6 X 10 mol’ s’, derived from the result
20
A solution of hCG was added Into the buffer solutIon at arrow A and was exchanged with a new buffer solution at B. The electrode was treated with an HCI
solution at C, the potential meastrement was reetmed at 0, andan hCGsolution
was added. The concentration of hCO was 3.3 mg/I. In the measwlng cell
(2)
where U1 is the initial potential value before the addition, Uf
is the potential expected at infinite reaction time, and a is the
product of the initial concentration of antigen, [B]0, and the
reaction rate constant, k. With this equation the rate constant
can be derived from analysis of the observed potential
curve.
We examined the effect of various speeds of rotation of the
stirrer on the potential behavior of the anti-hCG electrode,
using a constant concentration of hCG in the solution. The
resulting values of the rate constant obtained from equation
2, shown in Figure 3, indicate that the rate constant increases
gradually with speed of rotation, approaching a steady value
of k = 1.62 X 10 mol’ s’ at about 200 rpm.
For further understanding of the reaction rates obtained,
we analyzed the reaction by assuming that it is diffusion
controlled in the unstirred layer. The rate of the diffusional
transportation of the reactants through the unstirred layer
is expressed by
10
100
200
300
400
500
Rotating speed (rpm)
Fig. 3. The rate constant, k,
speed
as
a
function
The concentration of hCG added was fixed at 4.1 mg/I.
of
stIrrer rotatIon
(a)
>
‘a
3.0
2.0
(b
1.0
(C)
0
4
5
6
7
8
9
10
pH
time (mm)
FIg. 4. Response curves of the anti-hCG electrode on addition
of hCGto gIve concentrations of 12,5 (a), 6.25(b), and 4.1 (c)
FIg. 6. pH dependence of the potentIal change, U, - U1,of an
antl-hCG electrode caused by hCG, measured in 50 mmol/L
acetate buffer solutions (for pH <6) or In 50 mmol/L barbiturate
buffer solutions (for pH >6), at a stirrer speed of 333 rpm
mg/L
in solution has been reported for comparison with our data.
The rate constant for ovalbumin with its antibody was reported to be of the order of 10#{176}
mo11 s in solution at room
temperature
(19,20), which is two orders higher than our results. The reaction rate for the trypsin-aprotinin
reaction
derived from our potentiometric
experiment is 10-fold lower
than that measured in solution (12,13,21,22). The difference
in rate constants
between
ours and those taking place with
both reactants in solution is possibly due to the change of
steric configuration of biological substances caused by the
chemical modification.
The shift in potential, U1
seems independent
of the
concentration of added hCG and is almost the same for the
anti-hCG electrodes prepared in the same batch, but it varies
by from 1 to 8 mV for electrodes prepared in different batches.
This result suggests that the surface densities of anti-hCG
fixed on the electrodes differ for electrodes in different
batches. The magnitude, Uf U1, of the anti-hCG electrode
in the same batch is plotted as a function of pH in Figure 6,
where the Uf
U value was derived from equation 2, assuming first-order reaction. It has a maximum at about pH
= 8.7,and becomes zero below pH 5.The decline in U1
U
from pH 8.7 in either direction of pH may be explained, at
least in part, as due to the decrease in the equilibrium constant
for the antigen-antibody
reaction at low and high pH regions.
-
U1,
-
-
-
3.0
2.0
.
U,
I,’
1.0
0
0
2
4
6
8
10
(hg/mi)
FIg. 5. Relation of reaction rate, a, to concentration of hCG
hCG concentration
Because it is a common clinical practice to determine
pregnancy by detecting hCG in urine, we studied the response
of the anti-hCG electrode on adding 0.5 mL of urine to 6 mL
of buffer. The measurement was carried out for two samples
from pregnant women, three from non-pregnant
normal
women, and five from men. All samples, not only those from
pregnant women but also from normal women as well as men,
showed 2 to 4 mV shifts in potential, in the same direction.
When the sample solution was replaced with buffer solution
only, the potentials, except in the case of samples from pregnant women, returned to the initial value. In the case of
pregnant women, the potential was changed little by such
exchange of solutions. Thus, a response of the electrode to
hCG can be clearly distinguished from non-specific disturbances caused by other components of urine. Such nonspecific
responses may most probably be caused by physical adsorption of various components in urine. Filtration of urine samples did not eliminate these nonspecific responses. For comparison, we measured the concentrations of hCG in the same
specimens from the two pregnant women, with the semiquantitative latex-agglutination
test; both were found to be
between 300 and 600 mt. units/mL, 1 mt. unit corresponding
to about 0.1 gig. Consequently, the hCG concentrations in our
potential-measuring
vessel for the two samples from pregnant
women are estimated to be 2.3-4.6 mg/L.
The observed potential curves generally show no noise at
all, as seen from Figures 2 and 4. Therefore a very small change
in potential on addition of sample solutions can be recognized.
The smallest hCG concentration that can be detected in solutions of hCG only is determined mainly by the drift of the
original potential values, which may hide a very small shift on
sample addition. From several experiments with solutions of
hCG, we estimate the detection limit to be about 0.1
g/mL-i.e.,
1 mt unit/mL-about
the same sensitivity as for
the latex method. In the case of urine samples, the detection
limit for hCG cannot be clearly set with our present experimental procedure, because of the nonspecific disturbances by
the urine components, as mentioned above.
We hope that more sensitive and reliable electrodes will be
fabricated
as a result of searches for refined methods of
chemically modifying the electrodes. The use of high-purity
antibody may also be critically important for obtaining a good
electrode.
References
1. Giaever, I., The antibody-antigen
reaction: A visual
J. Immunol.
110, 1424-1426
observation.
(1973).
CLINICALCHEMISTRY,
Vol. 26,No. 11, 1980 1571
2. Janata, J., An immunoelectrode.
J. Am. Chem. Soc. 97,2914-2916
(1975).
3. Montgomery, M. R, Holtaman, J. L, and Leute, B.. K., Application
of electron spin resonance to determination of serum drug concentrations.Clin. Chem. 21, 1323-1328 (1975).
4. Katz, S. A., and Rechnitz, G. A., Direct potentiometric determination of urea after urease hydrolysis. Z. Anal. Chem. 196,248-251
13. Yaniamoto, N., Nagasawa, Y., Sawai, M., et al., Potentiometric
investigations of antigen-antibody and enzyme-enzyme
inhibitor
reactions using chemically modified metal electrodes.J. Immunol.
Methods 22, 309-317 (1978).
14. Okumura, H., Namba, S., and Matsushima, S., An unproved
purification of human chorionic gonadotropin. Endocri not. Jpn. 20,
(1963).
5. Guilbault, G. G., and Tarp, M., A specific enzyme electrode for
urea. Anal. Chim. Ada 73,335-365(1974).
6. Guilbault., G. G., Smith, B.. K., and Montalvo, J. G., Jr., Use of ion
selective electrodes in enzymic analysis. Cation electrodes for
deaminase enzyme systems. Anal. Chem. 41,600-605(1969).
7. Llenado, R. A., and Rechnitz, G. A., Automated serum urea determination using membrane electrodes. Anal. Chem. 46,1109-1112
15. Bahl, 0. P., Human chorionic gonadotropin. J. Biol. Chem. 244,
567-574(1969).
16. Porter, B..R., A chemical study of rabbit antiovalbumin. Biochem.
(1974).
8. Papastathopoulos, D. S., and Rechnitz, G. A., Enzymatic cholesterol determination using ion-selective membrane electrodes. Anal.
Chem. 47, 1792-1796 (1975).
9. Roes, J. W., Riseman, J. H., and Krueger, J. A., Potentiometric gas
sensing electrodes. Pure AppI. Chem. 36,473-487 (1973).
10. Hansen, E. H., and Ruzicka, J., Enzymatic analysis by means of
the air-gap electrode determination of urea in blood. Anal. Chim. Acta
72,353-364
(1974).
11. Aizawa, M., Kato, S., and Suzuki, S., Immunoresponsive membrane I. Membrane potential change associated with an immunochemical reaction between membrane-bound antigen and free antibody. J. Membrane Sci. 2, 125-132 (1977).
12. Yamamoto, N., Nagasawa, Y., Shuto, S., et al., The electrical
method of investigation of the antigen-antibody
and enzyme-enzyme
inhibitor reactions using chemically modified electrodes. Chem. Lett.
245-246 (1978).
1572
CLINICAL CHEMISTRY,
Vol.
26, No. 11, 1980
67-71 (1973).
J. 46,473-478 (1950).
17. Goldman, R., and Katchalski, E., Kinetic behavior of a two-enzyme membrane carrying out a consecutive set of reactions. J.
Theoret. Riot. 32,243-257 (1971).
18. Got, R., and Bourrillon, R., Nouvelle methode de purification de
Ia gonadotropine choriale humaine. Biochim. Biophys. Acta 42,
505-512 (1960).
19. Levison, S. A., Jancsi, A. N., and Dandliker, W. B., Temperature
effects on the kinetics of the primary antigen-antibody combination.
Biochem. Biophys. Res. Commun. 33,942-948(1968).
20. Levison, S. A., Kierszenbaum, F.,and Dandliker, W. B., Salt effects on antigen-antibody
kinetics. Biochemistry
9, 332-331
(1970).
21. Vincent, J. P., and Lazdunski, M., Trypsin-pancreatic trypsin
inhibitor association. Dynamics of the interaction and role of disulfide
bridges. Biochemistry
11, 2967-2977 (1972).
22. Finkenstadt, W. R., Hamid, M. A., Mattis, J. A., et al., Kinetics
and thermodynamics of the interaction of proteinases with protein
inhibitors. In Proteinase Inhibitors, H. Fritz, H. Tschesche, L. J.
Green, and E. Truscheit, Eds., Springer-Verlag, Berlin, 1974, pp
389-411.