capillary electrophoretic enzyme immunoassay with electrochemical detection for thyroxine
TRANSCRIPT
Capillary electrophoretic enzyme immunoassay withelectrochemical detection for thyroxine
Zhihui He1 and Wenrui Jin*
School of Chemistry and Chemical Engineering, Shandong University, Jinan 250100, PR China
Received 16 May 2002
Abstract
A capillary electrophoretic enzyme immunoassay with electrochemical detection (CE-EIA-ED) has been developed. In this
method, antigen (Ag) competes with horseradish peroxidase (HRP)-labeled antigen (HRP–Ag) for a limited number of antibody
(Ab) binding sites. The free HRP–Ag and the bound HRP–Ag–Ab complex are separated by capillary electrophoresis in a sepa-
ration capillary. Then they catalyze the oxidation of their enzyme substrate 3,30,5,50-tetramethylbenzide (TMB (reduced form)) with
H2O2 in a reaction capillary, which follows the separation capillary. The reaction product (TMB (oxidized form)) is amperomet-
rically determined using a carbon fiber microdisk bundle electrode at the outlet of the reaction capillary. Due to the amplification of
the enzyme, the concentration of TMB(Ox) is much higher than those of free HRP–Ag and the bound HRP–Ag–Ab complex.
Therefore, the limit of detection (LOD) of CE-EIA-ED is very low. The method has been used to determine thyroxine in human
serum. A concentration of LOD of 3:8� 10�9 mol/L, which corresponds to a mass LOD of 23.2 amol, was achieved.
� 2003 Elsevier Science (USA). All rights reserved.
Keywords: Capillary electrophoresis; Electrochemical detection; Immunoassay; Thyroxine
Capillary electrophoresis (CE)2 is a powerful tech-
nique for the separation of macromolecules such as
proteins and immunocomplex [1]. Several efforts have
been made to combine immunoassay with CE [2]. CE-
based immunoassay, called capillary electrophoretic
immunoassay (CEIA), offers several advantages such as
high selectivity, low reagent consumption, and short
incubation time over other conventional immunoassays[3]. CEIA can examine many components in a single
capillary because of its high separation efficiency. Fur-
thermore, this assay methodology can be simplified by
CE separation, in which many wash steps can be elim-
inated. CEIA has been applied to determination of hu-
man growth hormone [4], digoxin [5,6], insulin [7,8],
drugs [9–12], thyroxine [13], theophylline [14], immu-
noglobulin G [15], immunoglobulin A [16], bovine se-
rum albumin [17–20], and glucose-6-phosphate
dehydrogenase [21]. In CEIA, UV absorbance detection
[4,5] and laser-induced fluorescence detection [6–20]
have been used. However, the major disadvantage of theUV detection is the lack of sensitivity. Electrochemical
detection (ED) provides excellent sensitivity for the
small dimensions associated with CE, while offering a
high degree of selectivity toward electroactive species. In
this paper, a novel capillary electrophoretic enzyme
immunoassay with electrochemical detection (CE-EIA-
ED) is developed. In the assay, the immunoassay pro-
tocol was a competitive format, in which the antigen(Ag) competed with the horseradish peroxidase (HRP)-
labeled antigen (HRP–Ag) for a limited number of an-
tibody (Ab) binding sites. After equilibrium was estab-
lished, a small volume of incubate was injected into a
separation capillary, where the free HRP–Ag and the
bound antigen–antibody complex labeled by HRP
(HRP–complex) were separated by CE and catalyzed
Analytical Biochemistry 313 (2003) 34–40
www.elsevier.com/locate/yabio
ANALYTICAL
BIOCHEMISTRY
* Corresponding author. Fax: +86-531-8565167.
E-mail address: [email protected] (W. Jin).1 Present address: Technical Center of Changde Cigarette Factory,
Changde 415000, China.2 Abbreviations used: CE, capillary electrophoresis; CE-EIA-ED,
capillary electrophoretic enzyme immunoassay with electrochemical
detection; CEIA, capillary electrophoretic immunoassay; ED, electro-
chemical detection; HRP, horseradish peroxidase; SCE, saturated
calomel electrode; T4, thyroxine (3,5,30; 50-tetraiodo-LL-thyronine);
TMB, 3,3,5,5-tetramethylbenzidine; TMB(Ox), oxidized form of
TMB; TMB(Red), reduced form of TMB.
0003-2697/03/$ - see front matter � 2003 Elsevier Science (USA). All rights reserved.
PII: S0003 -2697 (02 )00508-0
the oxidation of the substrate, the reduced form of3,3,5,5-tetramethylbenzidine (TMB(Red)) in the reac-
tion capillary, which followed the separation capillary.
The catalysis reaction is shown as follows [22,23]:
The enzymatic reaction product is the oxidized form
of 3,3,5,5-tetramethylbenzidine (TMB(Ox)). TMB(Ox)
could be reduced at the carbon fiber microdisk bundle
electrode according to the following scheme [23].
The free HRP–Ag and the bound HRP–complex
could be detected through measuring TMB(Ox) at the
outlet of the reaction capillary. Since the concentration
of TMB(Ox) was much higher than those of the free
HRP–Ag and the bound HRP–complex due to the en-
zyme amplification, the limit of detection (LOD) of CE-
EIA-ED should be very low.Thyroxine (3,5,30; 50-tetraiodo-LL-thyronine, T4) is the
primary active hormone synthesized within the follicular
cells of the thyroid gland. It exerts regulatory effects on
target tissues. Measurement of T4 levels in serum is a
standard and well-validated test of thyroid gland func-
tion [24]. Therefore, T4 as the model antigen was in-
vestigated in CE-EIA-ED.
Materials and methods
Materials
An active T4 assay kit (Lot 08280) which consisted of
T4 standards containing 0, 5, 15, 50, 150, and 500 lg/LT4 in human serum, T4 controls (containing low andhigh concentrations of T4 in human serum), T4 enzyme
conjugate (HRP–T4), and anti-T4 mouse monoclonal
antibody was obtained from Diagnostics Systems Lab-
oratories (Webster, TX, USA). The kit was stored at
4 �C. All kit reagents and specimens were brought to
room temperature before use. HRP (R.Z. �3.0, 250U/
mg) was obtained from Shanghai Lizhu Dongfeng
Biotechnology (Shanghai, China). TMB(Red) was pur-chased from Sigma (St. Louis, MO, USA). A stock
standard solution of TMB(Red) (0.020mol/L) was pre-
pared in double-distilled water and kept in a dark bottle
and 0.020mol/L citrate–phosphate buffer (pH 5.0) was
prepared by dissolving the appropriate amount of cit-
rate and disodium hydrogen phosphate in double-dis-
tilled water. All buffers were filtered through 0.45-lmcellulose acetate membrane filters (Shanghai YadongResin, Shanghai, China) before use. The substrate so-
lution consisted of 2:0� 10�4 mol/L TMB(Red) and
0.020mol/L citrate–phosphate buffer (pH 5.0). The
running buffer consisted of 2:0� 10�3mol=L H2O2 and
5:0� 10�3 mol/L citrate–phosphate buffer (pH 5.0).
TMB(Red) or H2O2 was added to solutions just before
the measurement. The running buffer was renewed every
run. All buffers and solutions were stored at 4 �C. Otherreagents were of analytical grade and purchased from
standard reagent suppliers. All solutions were prepared
with double-distilled water and in disposable plastic
ware using disposable pipette tips. All disposable plas-
ticwares and disposable micropipette tips used in the
assay were autoclaved prior to use in order to denature
any contaminants.
CE-EIA-ED system
CE and ED in CE-EIA-ED used in this work were
similar to our previous description [25]. Briefly, a re-
versible high-voltage power supply (Model 9323HVPS,
Beijing Institute of New Technology, Beijing, China)
provided a variable voltage of 0–30 kV across the sep-
aration capillary with its outlet at ground potential. Thepoly(vinyl alcohol)-coated capillaries (50 lm ID, 375 lmOD) were purchased from Hewlett–Packard Instru-
ments. Pieces of 15 and 5 cm were used as separation
and reaction capillaries, respectively. ED at a constant
potential was also performed with the electrochemical
analyzer (Model CHI800, CH Instrument, Austin, TX,
USA). The detector and the catalysis reactor were
housed in a Faraday cage to minimize the interferencefrom noise of external sources. ED was carried out with
a three-electrode system. It consisted of a carbon fiber
microdisk bundle electrode as the working electrode, an
SCE used as the reference electrode, and a coiled Pt wire
(0.3mm diameter, 5 cm in length) placed at the bottom
of the cell as the auxiliary electrode. The arrangement of
the electrochemical detection cell has been illustrated in
detail [25]. The carbon fiber microdisk bundle electrodesused here were described previously [26]. Before use, all
carbon fiber microdisk bundle electrodes were cleaned in
alcohol and washed with double-distilled water for 5min
by an ultrasonicator. During electrophoresis, the elec-
trodes can be directly washed with alcohol and water in
the detection cell. Samples were injected hydrodynami-
cally.
Z. He, W. Jin / Analytical Biochemistry 313 (2003) 34–40 35
The CE-EIA-ED system is illustrated in Fig. 1. Itconsisted of six main parts, a high-voltage power supply
(1), a running buffer reservoir (2), a liquid pressure
buffer reservoir (3), a liquid pressure substrate reservoir
(4), a Plexiglas catalysis reactor (5), and an electro-
chemical detector (6). In the system, the cylindrical
running buffer reservoir (2) (12mm diameter and 20mm
in depth) was made from Plexiglas with a rubber cover
(9). There was metal tubing (7) in the reservoir wall. Themetal tubing linked up with the running buffer reservoir
(2) and the liquid pressure buffer reservoir (3) through a
plastic hose. On the hose there was a switch (8) to
control the flow from the liquid pressure buffer reser-
voir. In this system a metal needle (13) (400 lm ID,
680 lm OD) of a syringe with a hole (12) in the middle
passes through the catalysis reactor (5). Both the
poly(vinyl alcohol)-coated separation capillary (10) andthe poly(vinyl alcohol)-coated reaction capillary (11)
were inserted in the needle with a gap (about 10 lm)
between them. The separation capillary was connected
to the running buffer reservoir through the rubber cover.
The enzyme substrate (TMB(Red)) solution in the liquid
pressure substrate reservoir (4) was introduced into the
reaction capillary (11) through the gap by means of the
liquid pressure. There was another switch (80) to controlthe flow from the liquid pressure substrate reservoir (4).
A platinum wire (14) served as the grounded electrode in
contact with the substrate solution for the high potential
drop across the separation capillary. Unless noted
otherwise, the applied separation high voltage was
10 kV. In the system, all the joints could be fixed with
epoxy adhesive.
Immunoassay procedure
The immunoassay protocol was a direct competitive
format; 25 lL of the T4 standard or control, 2 lL of
HRP–T4, and 100 lL of anti-T4 mouse monoclonalantibody were added to a 0.6-mL microcentrifuge tube.
The solution was incubated for 30min at room tem-
perature and diluted to 250 lL by the running buffer.
Before injection, the levels of the solutions in the sample
vial (unshown in Fig. 1), in the running buffer reservoir
(2), and in the catalysis reactor (5) were kept at the same
height. The liquid pressure buffer reservoir (3) and the
liquid pressure substrate reservoir (4) were placed 40 cmover the running buffer reservoir and the catalysis re-
actor, respectively. The dynamical injection process was
as follows: first, the liquid pressure substrate reservoir
(4) was lowered keeping the solution level at the same
height as that of the catalysis reactor (5). The switch (80)of the liquid pressure substrate reservoir (4) was turned
on. The switch (8) of the liquid pressure buffer reservoir
(3) was turned off. Then hydrodynamic injection wasperformed by inserting the inlet of the separation cap-
illary in to the sample vial and raising the vial 6 cm in
height for 20 s. After injection, the separation capillary
was manipulated down, out of the sample vial, and then
immersed in the running buffer solution in the running
buffer reservoir (2). After that, the cover (9) of the
running buffer reservoir was sealed. The switch (8) of the
liquid pressure buffer reservoir was turned on, and thenthe liquid pressure substrate reservoir was raised 40 cm.
Finally, the separation high voltage was applied across
the separation capillary, the detection potential was
applied at the working electrode, and the electrophero-
gram was recorded. During the electrophoresis, the
same liquid pressure from the liquid pressure buffer
reservoir and the liquid pressure substrate reservoir was
kept to prevent a distortion of the flat electroosmoticflow profile in the separation capillary. The reaction
time could be controlled by the liquid pressure.
In the electrochemical detection, the working micro-
disk bundle electrode was cemented onto a microscope
Fig. 1. Overview of the CE-EIA-ED system. (1) High-voltage power; (2) running buffer reservoir; (3) liquid pressure buffer reservoir; (4) liquid
pressure substrate reservoir; (5) catalysis reactor; (6) electrochemical detector; (7) metal tubing; (8) and (80) switch; (9) and (90) rubber cover; (10)separation capillary; (11) reaction capillary; (12) hole; (13) syringe needle; (14) Pt wire.
36 Z. He, W. Jin / Analytical Biochemistry 313 (2003) 34–40
slide, which was placed over a laboratory-made xyz-micromanipulator and glued in place in such a way that
the microdisk end protruded from the edge of the slide.
The position of the microdisk bundle electrode was
adjusted (under a microscope) against the outlet of the
reaction capillary so that the electrode and the capillary
were in contact. This arrangement allowed easy removal
and realignment of both the capillary and the electrode.
All potentials were measured against SCE.
Results and discussion
Electrochemical behavior of TMB(Red) and TMB(Ox)
The electrochemical behavior of TMB(Red) and
TMB(Ox) on a glassy carbon electrode in 5:0� 10�2
mol=L Na2HPO4–2:5� 10�2 mol/L citrate (pH 5.0) has
been studied [23]. Two-electron redox behavior was
observed for TMB(Red) with oxidation peaks at 250
and 400mV. Its product, TMB(Ox) can be reduced on
the glassy carbon electrodes at 100mV. We reinvesti-
gated the electrochemical behavior of TMB(Red) and
TMB(Ox) at a carbon fiber microdisk bundle electrode
in the buffer (pH 5.0) by cyclic voltammetry. The cyclicvoltammogram is shown in Fig. 2. There are two oxi-
dation peaks at 290 and 520mV on the anodic branch
corresponding to TMB(Red) and two reduction peaks at
245 and 450mV on the cathodic branch corresponding
to the product of the oxidation of TMB(Red),
TMB(Ox). This means that TMB(Red) can be oxidized
when a potential higher than 150mV is applied and
TMB(Ox) can be reduced when a potential below500mV is applied. However, if the detection potentials
between 500 and 150mV are used, the oxidation of
TMB(Red) and the reduction of TMB(Ox) will simul-
taneously proceed. When both TMB(Red) and
TMB(Ox) are simultaneously present in the solution, the
reduction current of TMB(Ox) will be counteracted bythe oxidation current of TMB(Red). When the detection
potentials below 150mV are used, no oxidation of
TMB(Red) occurs. This means that in this case only the
reduction of TMB(Ox) is responsible for the detection
current, i.e., the detection current is corresponding only
to the concentration of TMB(Ox).
Optimization of CE-EIA-ED
Since HRP on both HRP–T4 and HRP–T4–anti-T4
could catalyze the enzyme substrate TMB(Red) in CE-
EIA-ED, HRP was used as the model compound for the
optimization of the CE-EIA-ED. In CE-EIA-ED, the
enzymatic reaction of TMB(Red) with H2O2 proceeds in
the reaction capillary of the catalysis reactor. The so-
lution eluting from the reaction capillary contains theproduct of the enzymatic reaction, TMB(Ox), the un-
reacted TMB(Red), and H2O2. Thus, TMB(Ox) with
low concentration at the outlet of the reaction capillary
is measured in the presence of TMB(Red) with high
concentration. Therefore, a detection potential suitable
to eliminate the oxidation current of TMB(Red) should
be selected. From the electrochemical behavior of
TMB(Red) and TMB(Ox) mentioned above, potentialslower than 150mV are suggested. At these potentials,
TMB(Ox) can be reduced and TMB(Red) cannot be
oxidized. Fig. 3 shows the relationship between the
electric charge detected (the area of the electrophoretic
peak), q, and the applied potential, Ed. When Ed is be-
tween 300 and 100mV, q is smaller and increases rapidly
with decreasing Ed. When Ed < 100 mV, q maintains a
constant value. In our experiments, the detection po-tential of 100mV was selected, because no oxidation of
Fig. 2. Voltammogram of 5:0� 10�4 mol/L TMB(Red) at a carbon
fiber bundle electrode in 5:0� 10�2 mol/L Na2HPO4–2:5� 10�2 mol/L
citrate (pH 5.0) in the presence of 1:0� 10�3 mol/L H2O2. Scan rate,
50mV/s.
Fig. 3. Relationship between the detected electric charge detected, q,
and the applied potential, Ed; running buffer, 2:0� 10�3 mol/L H2O2
in 2:6� 10�3 mol/L Na2HPO4–1:2� 10�3 mol/L citrate (pH 5.0);
substrate solution, 2:0� 10�4 mol/L TMB(Red) in 1:0� 10�2 mol/L
Na2HPO4–4:8� 10�3 mol/L citrate (pH 5.0); 0.5U/mL HRP; separa-
tion capillary, 30cm� 50lm ID, reaction capillary, 5cm� 50lm ID;
h, 40 cm; hydrodynamic injection, 6 cm for 20 s; electric field strengths,
667V/cm.
Z. He, W. Jin / Analytical Biochemistry 313 (2003) 34–40 37
TMB(Red) occurred and the current background wasvery low at this potential.
The electric charge detected, q, the migration time,tm,and the width at the half-peak, W1=2 on the electro-
pherograms and the number of theoretical plates, N, at
different liquid heights, h, in the system are listed in
Table 1. When h < 40 cm, q increases rapidly with in-
creasing h, which indicates that the increase of the so-
lution volume into the reaction capillary plays adominant role, though the reaction time is shortened.
When h > 40 cm, q decreases with increasing h. In this
case, the increase of the reaction time plays a dominant
role. q has a maximum at h ¼ 40 cm. tm decreases slowly
with increasing h. N and W1=2 are almost constants with
increasing h. Therefore, 40 cm was selected as the opti-
mum liquid height in our experiments.
Figs. 4A and B show the relationship between q andthe concentration of TMB(Red), CTMBðRedÞ, or the con-
centration of H2O2, CH2O2, respectively. When CTMBðRedÞ
< 4:0� 10�5 mol/L, q increases rapidly with increasing
of CTMBðRedÞ. When CTMBðRedÞ > 4:0� 10�5 mol/L, q is
almost constant, which indicates a substrate saturation
for HRP. When CH2O2< 1:0� 10�3 mol/L, q increases
rapidly with increasing CH2O2. When CH2O2
is between 1.0
and 3:0� 10�3 mol/L, q is almost constant, which indi-cates the saturation of H2O2 for HRP. However, when
CH2O2> 3:0� 10�3 mol/L, q decreases with increasing
CH2O2. This is because the activity of HRP is less sensitive
to excess H2O2 [27]. The results indicate that the maxi-
mum q was obtained using 2:0� 10�4 mol/L TMB(Red)
and 2:0� 10�3 mol/L H2O2. Therefore, these concen-
trations were used for determination of T4.
The electric field strength, E, exerts an influence ontm [28]. The effects of E on tm, q, W1=2, and N are listed
in Table 2. From Table 2, it can be seen that tm, q, and
W1=2 decrease with increasing E. N is almost constant.
q has the maximum at E ¼ 667V/cm. Therefore, E at
667V/cm was used.
CE-EIA-ED for T4
Assay of T4 was performed in the competitive format.
After the competition immunoreaction, the injected so-
lution contained T4, HRP–T4, T4–anti-T4 (complex of
T4 and anti-T4), and HRP–T4–anti-T4, which were re-
solved in the separation capillary. Both HRP–T4 and
HRP–T4–anti-T4 from the separation capillary catalyzed
the oxidation of TMB(Red) with H2O2 in the reaction
capillary, and the reaction product TMB(Ox) was de-
tected at the outlet of the reaction capillary. So two peaks
corresponding to HRP–T4–anti-T4 and HRP–T4 shouldappear on the electropherogram. Typical electrophero-
grams of the solutions containing HRP–T4 and anti-T4
without and with T4 are shown in Fig. 5. The lower curve
(curve 1), which consists of two peaks (peak 1 and peak 2),
Table 1
The values of q, tm;W1=2 and N at different liquid heights
h (cm) q (nC) tm (s) W1=2 (s) N � 103
28 4.5 618 20.3 5.1
32 6.0 604 19.8 5.2
36 11.1 595 20.5 4.7
40 15.1 589 20.1 4.8
45 9.3 584 19.5 5.0
50 7.2 580 20.4 4.5
Conditions: Ed, 100mV. Other conditions as in Fig. 3.
Fig. 4. (A) Relationship between the electric charge detected, q, and the
concentration of TMB(Red), CTMBðRedÞ, 2:0� 10�3 mol/L H2O2; (B)
relationship between the electric charge detected, q, and the concen-
tration of H2O2, CH2O2. 2:0� 10�4 mol/L TMB(Red); Ed, 100mV.
Other conditions as in Fig. 3.
Table 2
The values of q, tm;W1=2 and N at different electric field strengths
E (V/cm) q (nC) tm (s) W1=2 (s) N � 103
667 22.0 324 16.5 2.1
1000 20.1 310 15.6 2.2
1200 18.8 302 15.0 2.2
1333 17.2 295 14.7 2.2
Conditions: separation capillary, 15cm� 50lm ID; Ed, 100mV.
Other conditions as in Fig. 3.
38 Z. He, W. Jin / Analytical Biochemistry 313 (2003) 34–40
is from a solution containing only HRP–T4 and anti-T4.
The upper curve (curve 2) is from a solution containingHRP–T4 and anti-T4 with 50 lg/L T4. Compared with
curve 1, q of peak 1 decreases and q of peak 2 increases,
and the migration times of peak 1 and peak 2 are almost
unchanged. According to the principle of the competitive
assay, peak 1 and peak 2 can be identified as the peak of
HRP–T4–anti-T4 and the peak of HRP–T4, respectively.
Calibration curves based on the peaks of HRP–T4
and HRP–T4–anti-T4 for determination of T4 are
shown in Fig. 6, curves 1 and 2, respectively. The pointsin the curve represent the average q from three runs. The
plot demonstrates that increasing standard T4 causes
the expected increase of q of HRP–T4 and the expected
decrease of q of HRP–T4–anti-T4. Quantification could
be performed by the both calibration curves. The rela-
tive standard deviations (RSDs) of q for a series of 10
injections of the solution containing HRP–T4 and anti-
T4 without T4 were 6.8% for the peak of HRP–T4–anti-T4 and 3.1% for the peak of HRP–T4. The LOD for the
competitive assay could be calculated by using the mean
determination of the peak of HRP–T4 or HRP–T4–anti-
T4 for the zero-dose T4. Using the peak of HRP–T4,
LOD, defined as the mean q of HRP–T4 of the zero-
dose T4 calibrator plus three times its standard devia-
tion calculated from 10 trials of the calibrator, was
3.0 lg/L (3:8� 10�9 mol/L). The lowest measured T4calibrator was 5 lg/L. Using the peak of HRP–T4–anti-
T4, LOD, defined as the mean q of HRP–T4–anti-T4 of
the zero-dose T4 calibrator minus three times its stan-
dard deviation calculated from 10 trials of the calibra-
tor, was 25 lg/L (3:2� 10�8 mol/L). Because the peak of
HRP–T4 has low LOD and low RSD, q of HRP–T4 was
used for quantitation in the assay.
In this study, the injection volume calculated was6.1 nL according to the Hagen–Poiseuille equation for
the hydrodynamic injection with 6 cm height for 20 s.
Therefore, a mass LOD of 18 fg (23.2 amol) was ob-
tained. The linear range extended from the LOD up to
50 lg/L. Least-squares treatment of these data yielded a
slope of 0.251 nC L/g, an intercept of 9.88 nC, and a
correlation coefficient of 0.9997. To verify the method,
after being diluted 10 times two T4 controls (containing25� 7:5 and 100� 30lg=L in serum) provided by the
kit were detected. The results are shown in Table 3. The
concentrations of T4 controls determined were 31 and
112 lg/L, respectively, which agreed with the stated
values of the kit. The recovery of the method was be-
tween 90 and 105% for the detection of T4 controls.
Conclusions
The described CE-EIA-ED is a useful new method
with high selectivity, low LOD, and low sample con-
sumption for biological substances. In our work, a
Fig. 6. Calibration curves based on the peak of HRP–T4 (curve 1) and
the peak of HRP–T4–anti-T4 (curve 2) for determination of T4 by
competitive assay. Separation capillary, 15cm� 50lm ID; Ed,
100mV. Other conditions as in Fig. 3.
Fig. 5. Electropherograms of the solutions containing HRP–T4 and
anti-T4 without T4 and with 50lg/L T4. Separation capillary,
15cm� 50lm ID; Ed, 100mV. Other conditions as in Fig. 3.
Table 3
Results detected and recovery of T4
Sample Determined value (lg/L) Average value (lg/L) Added value (lg/L) Observed value (lg/L) Recovery (%)
I 32 31 50 79 94
30 150 165 90
II 114 112 150 271 105
109 250 335 90
Separation capillary, 15cm� 50lm ID; Ed, 100mV. Other conditions as in Fig. 3.
Z. He, W. Jin / Analytical Biochemistry 313 (2003) 34–40 39
concentration LOD of 3:8� 10�9 mol/L and a massLOD of 23.2 amol was achieved for T4 in serum sam-
ples. Thus, CE-EIA-ED is potentially applicable to de-
termine attomolar levels of biological substances. Since
many of the enzyme-linked immunosorbent assay
(ELISA) kits with labeled HRP on antigen or antibody
are commercially available, and TMB(Red) is used as
the enzyme substrate in the ELISA kits, CE-EIA-ED
can be always carried out based on the reduction ofTMB(Ox) on the carbon fiber microdisk bundle elec-
trode. In addition, this method is useful where the HRP
enzyme label is available and a fluorescent label is not.
We think that CE-EIA-ED will become a useful tool in
immunological assays.
Acknowledgments
This project was supported by the National Natural
Science Foundation of China and the State Key Labo-
ratory of Electroanalytical Chemistry, Changchun Insti-
tute of Applied Chemistry, Chinese Academy of Sciences.
References
[1] P.D. Grossman, J.C. Colburn, H.H. Lauer, R.G. Nielsen, R.M.
Riggin, G.S. Sittampalam, E.C. Rickard, Application of free-
solution capillary electrophoresis to the analytical scale separation
of proteins and peptides, Anal. Chem. 61 (1989) 1186–1194.
[2] J.J. Bao, Capillary electrophoretic immunoassays, J. Chromatogr.
B 699 (1997) 463–480.
[3] D. Schmalzing, W. Nashabeh, X. Yao, R. Mhatre, F.E. Regnier,
N.B. Afeyan, M. Fuchs, Capillay electrophoretic-based immuno-
assay for cortisol in serum, Anal. Chem. 67 (1995) 606–612.
[4] R.G. Nielsen, E.C. Richard, P.F. Santa, D.A. Sharknas, G.S.
Sittampalam, Separation of antibody–antigen complexes by
capillary zone electrophoresis, isoelectric focusing and high-
performance size-exclusion chromatography, J. Chromatogr. 539
(1991) 177–185.
[5] X. Liu, Y. Xu, M.P.C. Ip, Capillary electrophoretic enzyme
immunoassay for digoxin in human serum, Anal. Chem. 67 (1995)
3211–3218.
[6] F.-T.A. Chen, S.L. Pentoney Jr., Characterization of digoxigenin-
labeled B-phycoerythrin by capillary electrophoresis with laser-
induced fluorescence, J. Chromatogr. 680 (1994) 425–430.
[7] N.M. Schultz, R.T. Kennedy, Rapid immunoassays using capil-
lary electrophoresis with fluorescence detection, Anal. Chem. 65
(1993) 3161–3165.
[8] N.M. Schultz, L. Huang, R.T. Kennedy, Capillary electrophore-
sis-based immunoassay to determine insulin content and insulin
secretion from single islets of langerhans, Anal. Chem. 67 (1995)
924–929.
[9] L. Steinmann, J. Caslavaska, W. Thormann, Feasibility study of a
drug immunoassay based on micellar electrokinetic capillary
chromatography with laser-induced fluorescence detection: deter-
mination of theophylline in serum, Electrophoresis 16 (1995)
1912–1916.
[10] F.-T.A. Chen, R.A. Evangelista, Feasibility studies for simulta-
neous immunochemical multianalyte drug assay by capillary
electrophoresis with laser-induced fluorescence, Clin. Chim. 40
(1994) 1819–1822.
[11] R.A. Evangelista, F.-T.A. Chen, Analysis of structural specificity
in antibody–antigen reactions by capillary electrophoresis with
laser-induced fluorescence detection, J. Chromatogr. 680 (1994)
587–591.
[12] J. Caslavska, D. Allemann, W. Hormann, Analysis of urinary
drugs of abuse by a multianalyte capillary electrophoretic immu-
noassay, J. Chromatogr. A 838 (1999) 197–211.
[13] D. Schmalzing, L.B. Koutny, T.A. Taylor, W. Nashabeh, M.
Fuchs, Immunoassay for thyroxine (T4) in serum using capillary
electrophoresis and micromachined devices, J. Chromatogr. B 697
(1997) 175–180.
[14] N.H. Chiem, D.J. Harrsion, Microchip-based capillary electro-
phoresis for immunoassay: analysis of monoclonal antibodies and
theophyline, Anal. Chem. 69 (1997) 373–378.
[15] O.-W Reif, R. Lausch, T. Scheper, R. Freitag, Fluorescein
isothiocyanate-labeled protein G as an affinity ligand in affinity/
immunocapillary electrophoresis with fluorescence detection,
Anal. Chem. 66 (1994) 4027–4033.
[16] F.-T.A. Chen, Characterization of charge-modified and fluores-
cein-labeled antibody by capillary electrophoresis using laser-
induced fluorescence, J. Chromatogr. 680 (1994) 419–423.
[17] J.P. Ou, Q.G. Wang, T.M. Cheung, S.T.H. Chan, W.S.B. Yeung,
Use of capillary electrophoresis-based competitive immunoassay
for a large molecule, J. Chromatogr. B 727 (1999) 63–71.
[18] Q.G. Wang, G.A. Luo, Y.M. Wang, W.S.B. Yeung, Capillary
electrophoresis based immunoassay for monoclonal antibody with
diode laser induced fluorescence detection, Anal. Lett. 33 (2000)
589–602.
[19] J.P. Ou, S.T.H. Chan, W.S.B. Yeung, Separation of bovine serum
albumin and its monoclonal antibody from their immunocom-
plexes by sodium dodecyl sulfate-capillary gel electrophoresis and
its application in capillary electrophoresis-based immunoassay, J.
Chromatogr. B 731 (1999) 389–394.
[20] Q.G. Wang, G.A. Lou, J.P. Ou, W.S.B. Yeung, Noncompetitive
immunoassay using protein G affinity capillary chromatography
and capillary electrophoresis with laser-induced fluorescence
detection, J. Chromatogr. A 848 (1999) 139–148.
[21] Z. Rosenzwing, E.S. Yeung, Laser-based particle-counting micro-
immunoassay for the analysis of single human erythrocytes, Anal.
Chem. 66 (1994) 1771–1776.
[22] G.G. Guilbault, in: Handbook of Enzymatic Methods of Anal-
ysis, Marcel Dekker Inc., New York and Basel, 1976, p. 147.
[23] G. Volpe, D. Compangnone, R. Draisci, G. Palleschi, 3,30,5,50-
tetramethylbenzidine as electrochemical substrate for horseradish
peroxidase based enzyme immunoassay, Analyst 123 (1998) 1303–
1307.
[24] J.G. Gruhn, C.P. Barsano, Y.A. Kumar, The development of tests
of thyroid function, Pathol. Lab. Med. 111 (1987) 84–100.
[25] W. Jin, Q. Weng, J. Wu, Determination of bovine serum albumin
by capillary zone electrophoresis with end-column amperometric
detection at the carbon fiber microdisk array electrode, Anal.
Chim. Acta 342 (1997) 67–74.
[26] W. Jin, D. Yu, Q. Dong, X. Ye, Quantitation of caffeine by
capillary zone electrophoresis with end-column amperometric
detection at a carbon microdisk array electrode, J. Chromatogr.
Sci. 38 (2000) 11–15.
[27] N. Hoshio, R. Nakajima, I. Yamazaki, The effect of polymeriza-
tion horseradish peroxidase activity in the presence of excess
H2O2: a background for a homogeneous enzyme immunoassay, J.
Biochem.-Tokyo 102 (1987) 785–791.
[28] J.W. Jorgenson, K.D. Lukacs, High-resolution separation based
on electrophoresis and electroosmosis, J. Chromatogr. 218 (1981)
209–216.
40 Z. He, W. Jin / Analytical Biochemistry 313 (2003) 34–40