| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
Departments of
1
Pathology and
2
Biochemistry and Molecular Biophysics, Medical College of Virginia, Virginia Commonwealth University, Richmond, VA 23298-0597.
3
Current address: Roche Diagnostics, 1080 US Highway 202,
Somerville, NJ 08876.
a Author for correspondence: Medical College of Virginia, Virginia Commonwealth University, P.O. Box 980286, Richmond, VA 23298-0286. Fax 804-828-0353; e-mail millerg{at}gems.vcu.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Biosensor measurements of specific analytes within complex sample matrices have utilized a variety of chemical indicator systems, including organic molecules, enzymes, antibodies, and receptors to isolate the desired analyte from the sample milieu with high specificity (1)(2)(3). The recognition event usually does not generate a measurable signal in situ and therefore requires coupling to a transduction system. A variety of electrochemical and optical transduction systems have been explored to quantitate various analytes. Optical sensors have the desirable attributes of immunity to electrical noise, no risk of electrical shock to the patient, and physical flexibility, including catheter placement, to reach remote locations.
Sensor systems with antibodies for molecular recognition can be applied to a large variety of analytes. Previously reported self-contained immunosensors have measured low-molecular-mass analytes. Bush and Rechnitz (4) used an electrochemical transducer with a membrane-encapsulated antibody indicator system to quantitate dinitrophenol. Anderson and Miller (5) and Hanbury et al. (6) described a continuously monitoring fiber-optic immunosensor with a fluorescence transduction system and a membrane-encapsulated competitive-binding antibody system to measure aqueous phenytoin and theophylline concentrations. Astles and Miller (7) reported the continuous measurement of phenytoin in serum and whole blood with this sensor design. Various other fiber-optic sensor designs have used antibody or enzyme transduction chemistry covalently attached to an optical fiber (8), attached to an optical component interrogated by the fiber (9), incorporated into a polymer matrix (10), or temporally released from polymer gels (11). Antibody-based sensors for measuring large protein analytes have not been self-contained and thus have not been suitable for continuous monitoring applications (12)(13).
One of the challenges for fiber-optic immunosensor construction is the development of a self-contained packaging design for the chemical indicator system that isolates the analyte of interest from the sample matrix. The design of reagent immobilization at the sensing area must maintain the functional activity of the recognition element and provide selective permeability for the desired analyte without excessively increasing sensor response time. Previous work from this laboratory reported a self-contained immunosensor for measurement of low-molecular-mass analytes (5)(6)(7).
We report the development of an antibody-based fiber-optic immunosensor to measure a protein analyte. Myoglobin was an excellent model protein analyte because its moderate molecular dimensions, 16 500 Da, allowed diffusion through immobilization media, and the heme group has a large molar absorptivity that could be exploited in a fluorescence energy transfer design.
Myoglobin and other cardiac proteins have clinical relevance because they are measurable at the time of admission to the emergency department after myocardial infarction (14). Recent advances in the treatment of acute coronary syndromes and assessment of thrombolytic therapy have demonstrated the importance of rapid assessment of patients presenting to the emergency department with chest pain (15). One can speculate that availability of a catheter-based sensor for two or three cardiac proteins could accelerate effective therapeutic intervention in this critical condition. Coagulation proteins during surgery represent another acute-care situation in which in vitro measurement might be useful.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Conventional fluorescence measurements of Cascade Blue were performed in a 1 x 1 cm quartz cuvette at room temperature with a spectrofluorometer (MPF-44A, Perkin-Elmer) with a 150 W xenon arc lamp source. An excitation wavelength of 375 nm and emission wavelength of 425 nm were used for Cascade Blue. Approximately equal width slits (<4-nm band pass) were used to achieve full-scale deflection on a chart recorder. All reagent additions and mixing by volume exchange with a transfer pipette were performed manually without removing the cuvette from the sample compartment.
fiber-optic fluorometer
Figure 1
shows a diagram of the fiber-optic fluorometer constructed in
this laboratory. All optical components were mounted on a 14.2 x
9.4 cm optical table (XI-23, Newport Research). Excitation radiation
was from a 75 W xenon arc lamp (75XBO75/2OF, Osram) mounted
horizontally within a f/4.5 elliptical reflector in an
air-tight, water-cooled housing (HH150, Photon Technology) with a
constant-current power supply (LPS 200X, Photon Technology) operated at
5 A. Excitation light was passed through a heat-absorbing glass filter;
a 370-nm, 6-nm band pass filter (PIO 370-F, Corion Corp.); and then
through a 2-mm hole in the central axis of a 66-mm off-axis parabolic
mirror (02POA015, Melles Griot) mounted in a three-plane adjustable
mount. Light passing through the mirror was collimated with a 50-mm
diameter, 76-mm focal length, fused-silica biconvex lens (01LDX145,
Melles Griot), and refocused on an optical fiber with a 25-mm diameter,
f/1 fused-silica biconvex lens (01LQB028, Melles Griot). The
lenses were arranged so that the bright spot of the lamp was at the
focal point of the first lens, with the light traversing ~40 mm
between the two lenses, after which the image of the lamp was refocused
on the proximal end of an optical fiber. A five-axis fiber-optic
positioner (FP-2, Newport) was used to align the optical fiber at the
focal point.
|
Excitation light passed through ~1 m of multimode step-index optical fiber (320-µm diameter core, 32.5-µm cladding, and 385-µm diameter total) with polyimide coating (Superguide G, Thermocoat, SFS320/385T, Fiberguide Industries). This fiber had negligible fluorescence at 425 nm when excited at 370 nm. The polyimide protective coating was removed from ~30 mm at each end of the optical fiber by charring with a butane burner, followed by cleaning with deionized water and acetone. Both ends of the optical fiber were cleaved at 90° to the longitudinal optical axis with a diamond edge cleaver (FK-LDF, York Technology), cleaned in 50 mL/L nitric acid for 15 min, soaked in a silanizing reagent (SL-2, Sigmacote, Sigma) for 5 min to block free hydroxyl groups, and rinsed with deionized water before use.
Fluorophore emission that entered the distal end of the waveguide was transmitted back through the fiber, collimated by the input lens, and reflected 90° and focused by the parabolic mirror through a 5-nm band pass entrance slit of a grating monochrometer (PTR Optics) set at 425 nm, and through a 425-nm, 5-nm band pass interference filter (XMM-425-F, Corion). The signal was detected with a photomultiplier tube (1P28B22, RCA) operated at -650 V dc (Stanford Research Systems) and measured with an analog display photon counter (126, Pacific Instruments). To prevent photodegradation of the reagents, incident light was blocked with a mechanical shutter that was opened only during data collection. Fluorometer components were aligned along the excitation optical axis with the incident arc lamp beam. A helium/neon laser was coupled temporarily to the distal end of the fiber to align the return fluorescence into the monochrometer slit.
A reference channel was constructed with 0.5 m of multimode step-index optical fiber (100-µm core, 10-µm cladding, and 120-µm total diameter) with polyimide coating (Superguide G, SFS100/100T, Fiberguide) placed 5 mm from the lamp housing leading to a photomultiplier tube (1P28B22, RCA) operated at -650 V dc (Pacific Instruments). The photomultiplier tube output was measured with a high-impedance digital voltmeter.
Fluorescence measurements of solution-phase reactions were made with the fiber-optic instrument by submerging the distal end of the bare fiber into test samples. The fluorescence was determined for each reagent system contained in a 15 x 75 mm vial with constant mechanical stirring.
reagents
All chemicals were reagent grade except where indicated. Distilled
deionized water was used throughout. Human cardiac myoglobin (98+%
pure by sodium dodecyl sulfate electrophoresis, Fortron Bioscience) was
diluted in pH 7.4 phosphate-buffered saline
(PBS1
; 10 mmol/L sodium phosphate and 150 mmol/L sodium chloride) and
stored at 4 °C until use.
Human hemoglobin was isolated from a hemolysate prepared from a fresh 100% hemoglobin A whole-blood sample collected in EDTA. The sample was centrifuged (1500g for 10 min), and the supernatant and buffy coat were discarded. A 2-mL cell pellet was washed twice with 4 mL of 9 g/L sodium chloride solution, and both supernatants were discarded. Erythrocytes were lysed by adding 1 mL of deionized water to the remaining pellet, followed by vortex-mixing and freezing at -70 °C overnight. The sample was thawed and centrifuged to pellet cellular debris, and the supernatant was transferred to a capped cryovial for storage at -20 °C. The final hemoglobin concentration was 2.3 mmol/L (148 g/L) (Cell-Dyn; Abbott Diagnostics). Hemolysate (9 mL) was diluted in 20.0 mL of PBS to give a final concentration of 1 mmol/L (66 mg/L).
purification of anti-myoglobin antibody and conjugation to cascade
blue
Five monoclonal IgG1 
myoglobin antibody
clones were evaluated for use in the sensor. Antibody clones were each
purified by Protein A chromatography (Immunopure IgG, Pierce) and
dialyzed in a 12 000-Da cutoff cellulose membrane against triplicate
changes of 1 L of PBS. The final postdialysis antibody concentration
was ~1.1 g/L as calculated from the molar absorptivity, 2.1 x
105 L mol-1 cm-1 at 280 nm
(17). Antibody purity was confirmed by observation of a
single electrophoresis band for each clone with both 7.0% homogeneous
and 8–25% gradient polyacrylamide gels with Coomassie Blue staining
(PhastGel, Pharmacia Biotech).
After purification, antibodies were conjugated to Cascade Blue (Molecular Probes). Cascade Blue acetylazide was obtained at 60% purity (by weight), and all molar ratios were adjusted accordingly. For the initial experiments to determine the optimum Cascade Blue:antibody molar ratio, Cascade Blue acetylazide (5 mg, 4.9 µmol) was dissolved in 500 µL of N,N-dimethylformamide (22,705-6, Aldrich) made anhydrous by drying over molecular sieves (20,860-4, 8–12 mesh, Aldrich) to give a final concentration of ~9.9 mmol/L. Aliquots of five antibody clones were diluted to ~0.73 µmol/L with PBS and dialyzed against 10 mmol/L pH 8.3 sodium bicarbonate buffer. Cascade Blue acetylazide was added at 27-, 135-, 675-, and 1350-fold molar excess (dye:antibody) to 1 mL of each antibody solution and incubated for 4 h at room temperature. Freshly prepared hydroxylamine (10 µL of 15 mol/L reagent in pH 8.3 bicarbonate buffer) was then added to each labeled antibody solution to stop the reaction and remove labile intermediates from tyrosine, serine, threonine, and histidine. The solution was incubated for an additional 30 min at ambient temperature. The reaction mixture containing labeled antibody was dialyzed in triplicate against 1 L of PBS over 48 h at 4 °C.
For all subsequent experiments, the labeling procedure was modified by increasing the Cascade Blue and antibody concentrations to reduce azide hydrolysis. Each antibody concentration was increased to 7.3 µmol/L and Cascade Blue acetylazide was increased to 16.5 mmol/L for a fluorophore:antibody molar ratio of 823. The reaction mixture was incubated for 4 h, after which hydroxyl-amine was added to stop the reaction, diluted to 1 mL total volume with PBS, and dialyzed as described above. Because Cascade Blue absorbs light at 280 nm, the final antibody concentration (~5 µmol/L) was calculated from total nitrogen content. Cascade Blue concentration was determined from its molar absorptivity, 28 000 L mol-1 cm-1 at 376 nm (18).
measurement of cascade blue antibody reaction with myoglobin
Typically, 10 µL of ~7.3 µmol/L Cascade Blue-labeled
antibody was added to 990 µL of PBS to give a 73 nmol/L solution. The
fluorescence was measured at 425 nm, and 10-µL aliquots of 4.0
µmol/L myoglobin were added. The fluorescence quench was measured
after at least 30 min of incubation after each myoglobin addition. For
interference evaluation, hemoglobin was substituted for myoglobin.
preparation of aminoethylacrylamide
Aminoethylacrylamide was synthesized for covalent attachment to
antibodies by a modified procedure for coupling diazonium salts to
proteins (19)(20). Ethylenediamine (15 mmol)
(24,072-9, Aldrich) was added to 50 mL of dioxane/water (500 mL/L),
adjusted to pH 8.5 with 3 mol/L HCl, and chilled to 4 °C. Acryoyl
chloride (16 mmol) (A2,410-9, Aldrich) was diluted in 30 mL of
chloroform that was prechilled to 4 °C. The acryoyl chloride
solution was added slowly (~2 mL/min) with a dropping funnel to the
ethylenediamine solution while the pH was maintained between 8 and 8.5
with 3 mol/L sodium hydroxide with constant stirring in an ice bath.
After 2 h, the aqueous phase was concentrated by vacuum
evaporation, and the remaining slurry was redissolved in deionized
water. Particulates were removed by centrifugation at 1000g
for 15 min, and the supernatant was lyophilized and collected. Recovery
of aminoethylacrylamide-HCl was ~10 mmol, giving >65% yield.
conjugation of anti-myoglobin antibody to aminoethylacrylamide
Cascade Blue-labeled antibody was dialyzed against 1 L of PBS
adjusted to pH 4.7. Antibody (1 mL, containing ~7 nmol) was reacted
at 4 °C with 10 µmol of aminoethylacrylamide-HCl, 50 µmol of
1-(3-dimethylaminopropyl)-3-ethylcarbodiimide (16,146-2, Aldrich), and
50 µmol of N-hydroxysuccimide (H 7377, Sigma). The pH of
the reaction mixture was readjusted to 4.5, and the solution was
incubated at 4 °C for 24 h, followed by dialysis in
12 000-Da-cutoff cellulose tubing against quadruplicate changes of 1 L
of PBS. The final antibody concentrations were assumed to be equivalent
to the starting material because there was negligible volume change
during dialysis.
The effect of the aminoethylacrylamide conjugation on antibody binding was determined for three clones on the basis of the fluorescence quench after the addition of myoglobin. Each clone was diluted to ~73 nmol/L by the addition of 10 µL of labeled antibody to a 1 x 1 cm quartz cuvette containing 990 µL of PBS. The fluorescence was measured in a conventional fluorometer after the addition of ten 10-µL aliquots of 4.0 µmol/L myoglobin, allowing at least 30 min between additions to reach equilibrium.
determination of nonspecific fluorescence quench
Monoclonal theophylline antibody (Beckman Instruments), bovine
serum albumin (Sigma), and two monoclonal myoglobin antibody clones
were each diluted to 1.1 g/L in PBS. Aliquots (300 µL) of each
protein solution and a PBS control were simultaneously labeled with 100
µL of 16.5 mmol/L Cascade Blue acetylazide as described above. The
solutions were dialyzed separately against triplicate changes of 1 L of
PBS. The decrease in fluorescence was measured after the addition of
10-µL aliquots of 4.0 µmol/L myoglobin to 990 µL of PBS
containing 10 µL of each Cascade Blue-labeled protein.
polyacrylamide gel preparation
Polyacrylamide gels were prepared from aqueous stock solutions of
acrylamide (14,866-0, electrophoresis grade, Aldrich) and
N,N'-methylenebisacrylamide (BIS;
14-832-6, electrophoresis grade, Aldrich). Where indicated,
aminoethylacrylamide conjugated to myoglobin antibody was substituted
for BIS to achieve the total desired net concentration of acrylamide
incorporated into the gel. Acrylamide stock solutions were stored at
4 °C and used within 2 weeks to limit hydrolysis of acrylic acid.
Ammonium persulfate (24,861-4, ACS grade, Aldrich) free radical
generator was prepared immediately before use as a 100 g/L solution in
deionized water. Polyacrylamide gels were prepared by the addition of
209 µL of stock acrylamide/BIS solution, 350 µL of deionized water,
625 µL of pH 7.4 PBS, and 9 µL of ammonium persulfate. The
solutions were deoxygenated with dry nitrogen for 30 s at room
temperature, followed by the addition of 3 µL of 6.63 mol/L
N,N,N',N'-tetramethylethylenediamine
(T2,250-0, Aldrich) to catalyze the reaction.
The optimal acrylamide/BIS-acrylamide gel composition was determined on the basis of differential diffusion of myoglobin and hemoglobin in 3-mm-deep Ouchterlony diffusion gels cast in Petri dishes. Wells (2-mm diameter) were made at four locations on the plate and filled with 2 µL of 640 mg/L myoglobin or hemoglobin. The radial diffusion of myoglobin and hemoglobin was measured after a 24-h incubation at room temperature.
sensor construction
Sensors were constructed with two glass capillary tubes
surrounding an optical fiber, as shown in Fig. 2
. A 45-mm-long, 5-µL capillary tube (Drummond Scientific) was
fixed with cyanoacrylate adhesive inside a 50-mm-long, 25-µL
capillary tube, leaving a 5-mm cavity at the distal end of the tubes.
The sensor cavity was immersed in a 5.3% polyacrylamide solution
containing ~7 nmol/L Cascade Blue–antibody (0.05 volume fraction)
immediately after the addition of tetramethylethylenediamine. The
distal end of a prepared optical fiber was inserted through the lumen 3
mm beyond the end of the inner capillary and fixed with silicone
adhesive at the proximate capillary end. The outer capillary provided a
reproducible template for the polyacrylamide gel and provided
protection against mechanical disruption of the fiber in the soft gel
matrix. Sensors were incubated at room temperature for 1 h to
ensure complete polymerization within the gel cavity. The sensors were
incubated overnight at 4 °C in PBS to rinse adsorbed antibodies from
the gel and then stored at 4 °C in PBS in a sealed tube until used.
|
measurement of sensor response
Sensors were operated at room temperature in a covered vial. Each
sensor was initially equilibrated in 500 mL of PBS for 5 h at room
temperature to ensure signal stability. Fluorescence measurements were
taken periodically and recorded vs time. Sensor response was determined
after typical additions of 5–10 µL of stock myoglobin or hemoglobin
to 1 mL of PBS to give the desired concentrations. The sensor
fluorescence was measured until maximum quench was obtained. The sensor
response was determined as the slope of fluorescence vs time and as the
total amplitude of fluorescence quench measured in counts per second
(cps). After maximum quench was achieved, sensors were transferred to
PBS or to a solution of unlabeled myoglobin antibody in PBS to effect a
zero concentration of free myoglobin in the solution.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
shows the absorbance and emission spectra for myoglobin and
Cascade Blue. The high molar absorptivity of myoglobin at 418 nm,
12.8 x 101
L mol-1
cm-1, and good spectral overlap with the emission spectrum
of Cascade Blue enabled effective nonradiative energy transfer when the
two molecules were geometrically proximate after antibody binding to
myoglobin. Antibody binding was monitored by the quench of Cascade Blue
fluorescence at 425 nm.
|
Optimal labeling conditions were critical in achieving the maximum
change in fluorescence signal per unit of myoglobin. Fig. 4
shows the fluorescence quench obtained after addition of
myoglobin to antibody labeled with 27- to 1350-fold molar excess
Cascade Blue. A molar ratio of 675 or greater gave maximum fluorescence
quench. No increase in maximum quench was obtained by labeling at
larger molar ratios, presumably because all Cascade Blue conjugation
sites with geometric positions capable of energy transfer were
occupied. Subsequent conjugations were performed at higher antibody
concentrations with 823-fold molar excess Cascade Blue. These
conditions increased the signal change per unit of myoglobin ~2-fold,
presumably because of improved fluorophore conjugation efficiency.
|
The optimal labeling condition for achieving the maximum fluorescence
response range of the chemical transduction system required empirical
determination for each antibody clone. Fig. 5
shows the total fluorescence quench at three myoglobin
concentrations for five antibody clones labeled with the same molar
ratios of Cascade Blue. Antibody G gave the greatest signal change over
the largest range of myoglobin concentrations.
|
Figure 6
shows the reversibility of the response of the sensor chemistry
in solution phase. When myoglobin was added to a labeled antibody
solution, it bound rapidly to antibody and quenched the Cascade Blue
fluorescence. As more myoglobin was added, additional quenching was
observed. At approximately equimolar quantities of antibody and
myoglobin, a stable signal was observed (90–100 min) and we added a
9-fold excess of unlabeled myoglobin antibody from a different clone.
The unlabeled antibody rapidly bound any free myoglobin in solution and
scavenged any myoglobin that dissociated from the labeled antibody.
Thus, the Cascade Blue:antibody:myoglobin dissociation reaction was
observed as an increase in fluorescence that remained stable for 1
h. The dissociation reaction occurred in ~10 min in solution phase.
|
Figure 7
shows nonspecific fluorescence change attributable to inner
filter absorption from solution phase of myoglobin after addition to
Cascade Blue-labeled theophylline monoclonal antibody, Cascade
Blue-labeled albumin, and Cascade Blue hydrolyzed in buffer. The inner
filter absorption did not influence the fluorescence quench for
specific binding of myoglobin to myoglobin antibody.
|
Hemoglobin was evaluated as a potential interferent in the myoglobin
measurement system. Fig. 8
shows the results for addition of hemoglobin to several Cascade
Blue-labeled myoglobin antibodies and Cascade Blue in buffer. All of
the myoglobin antibodies showed hemoglobin interference. The minimum
fluorescence response to hemoglobin was observed with antibody D, but
this antibody also had the poorest response to myoglobin (Fig. 5
).
Antibody G had the largest quench response to myoglobin and hemoglobin.
The fluorescence quench with Cascade Blue in buffer was minimal,
suggesting no inner filter effect. The quench observed with hemoglobin
appeared to represent cross-reactivity in the myoglobin–antibody
binding reaction. Consequently, the sensor design required a physical
separation barrier to restrict hemoglobin access from the antibody
reaction area.
|
Polyacrylamide gel was evaluated as a medium to immobilize Cascade
Blue-labeled antibody at the distal end of an optical fiber and to
restrict access of, and thus potential interference from,
large-molecular-mass matrix components such as hemoglobin. Table 1
shows the radial diffusion of myoglobin and hemoglobin in
various polyacrylamide gels. Gels prepared at <3.5% total acrylamide
did not polymerize to a consistency rigid enough for use in the sensor.
Gels prepared at acrylamide:BIS molar ratios of 75 and 100 gave the
maximum diffusion of myoglobin with limited diffusion of hemoglobin.
The 5.3% gel with an acrylamide:BIS molar ratio of 75 was used in all
subsequent experiments because the lower gel concentration should
enable faster diffusion of myoglobin to the sensing area.
|
A Cascade Blue–(antibody–aminoethylacrylamide) conjugate was
covalently incorporated into the polyacrylamide gel to eliminate the
possibility of Cascade Blue-labeled antibody leaching from the polymer
matrix. Table 2
shows the effect on the antibody binding after conjugation of
aminoethylacrylamide to Cascade Blue-labeled antibody. The solution
phase fluorescence quench was slightly decreased for
aminoethylacrylamide-labeled antibodies compared with
non-acrylamide-labeled Cascade Blue–antibody controls. However, the
labeling process did not substantially reduce the performance of the
detection chemistry.
|
myoglobin sensor performance
Table 3
presents sensor responses when antibodies at the same
concentrations were covalently attached to and entrapped in the
polyacrylamide matrix. The total fluorescence response of sensors
prepared with covalently attached antibody was less than that of the
entrapped design. The rate of signal change with antibody covalently
bound to the gel was about half the rate observed with the entrapped
design, suggesting decreased relative diffusion of myoglobin through
the gel with covalently attached antibody. In the entrapped
configuration, the mean total quench was about the same at both
myoglobin concentrations, but the rate of signal change was
concentration-dependent. These observations suggest that diffusion into
the gel matrix continued until all accessable antibody was bound to
myoglobin at equilibrium.
|
The total fluorescence for an intact sensor assembled with antibody labeled with 823-fold molar excess Cascade Blue was ~37 000 cps. The background signal for the fiber-optic fluorometer was ~15 000 cps for a bare fiber submerged in deionized water and 16 500 cps for a sensor constructed with polyacrylamide without Cascade Blue-labeled antibody. Two different sensors, prepared with antibodies C and G, had fluorescence signals at 24 h that were within 150 cps of the starting fluorescence. These results indicated that entrapped or polymerized Cascade Blue-labeled antibodies were equally retained in the 5.3% polyacrylamide gel matrix. Conjugation of antibody in the polyacrylamide gel reduced performance, was not necessary to prevent leaching, and was not investigated further.
Figure 9
shows a sensor's response to myoglobin and hemoglobin. The
sensor had a stable response with ±50 cps (1.5%) noise in buffer
between 0 and 80 min. Stabilization after the myoglobin additions was
also ±50 cps between 120–170 min and 300–420 min. No fluorescence
quench was observed with hemoglobin at a concentration of 93 nmol/L,
indicating the effectiveness of the polyacrylamide gel as a
permeability barrier to the larger molecular mass protein. The sensor
had a 200 cps response at 20 nmol/L (330 µg/L) myoglobin and an
additional 400 cps fluorescence decrease at 190 nmol/L (3.1 mg/L). A
reversible response was not observed when the sensor was removed from
the myoglobin solution and transferred to buffer containing 730 nmol/L
unlabeled myoglobin antibody (90-fold molar excess to that in the
sensor). This experiment was repeated with a different sensor that had
a 50 cps quench (barely discernible from the baseline noise) when
immersed in 5 nmol/L myoglobin and a similar 600 cps quench in 190
nmol/L myoglobin. This sensor also did not have a reversible response
when placed into a 0 concentration myoglobin solution. The limit of
detection for the sensor was ~5 nmol/L (83 µg/L), and incremental
fluorescence responses to incremental increases in myoglobin
concentration were observed.
|
In Fig. 9
the sensor response time to reach a stable signal was 20 min
at 20 nmol/L. At 190 nmol/L the response time was between 50 and 135
min but cannot be further defined because of the lack of observations
past 50 min. Three other sensors measured concentrations of 190–353
nmol/L with an average response time of 60 min (range 48–66 min).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The antibody was immobilized at the distal end of an optical fiber in a geometric configuration that provided a self-contained sensor unit. Polyacrylamide was selected as the immobilization medium because of its flexibility in creating different pore sizes to include or exclude molecules over a relatively large molecular-mass range, its polymerization at physiological pH to preserve the integrity of the antibody, and its low adsorption of proteins. The gel pore size was adjusted to retain the fluorescently labeled antibody (150 000 Da) in a reactive configuration, prevent the diffusion of the interferent hemoglobin (64 500 Da), and allow the diffusion of myoglobin (16 500 Da) to the sensing area with an usable response time. Hemoglobin was a good test molecule for high-molecular-mass interferents because its size is similar to albumin and because it cross-reacts with the antibody-based fluorescence transduction chemistry. Thus the sensor response provided a sensitive indicator of the effectiveness of the gel entrapment design to prevent interference from molecules similar in size to or larger than hemoglobin. Specific serum and blood interferents would need to be tested in further development of this design, but one would expect the physical barrier to hemoglobin to apply similarly to other proteins and to cells.
The thickness of the gel layer was selected to allow adequate labeled antibody for a measurable fluorescence signal while permitting diffusion of myoglobin analyte into the optically interrogated region of the sensor. Simple entrapment of antibody in the polyacrylamide effectively retained antibody in the sensor region and produced a stable fluorescence signal. Covalent attachment of antibody to the polyacrylamide matrix was not required and somewhat degraded the sensor response. However, covalent attachment is feasible and provides another means for further optimization of the gel and for extension to other analytes.
The fiber-optic sensor was stable in buffer and showed a concentration-dependent fluorescence response to myoglobin. The slower antibody response in a sensor compared with solution-phase reactions was likely a result of slower diffusion of the protein analyte through the polyacrylamide matrix. It appears that myoglobin diffused into the sensing region at a rate controlled by both the gel pore size and the concentration gradient from bulk solution into the gel. Because antibody in the gel did not functionally release myoglobin after binding, a nearly constant concentration gradient was maintained until myoglobin binding reached an equilibrium with available antibody binding sites.
The analytical lowest detection limit of the current sensor configuration was marginal for measurement of myoglobin near the clinical decision limit after a myocardial infarction (5 nmol/L; 83 µg/L) but consistently showed a measurable signal at 20 nmol/L (330 µg/L). The sensor stability was adequate for an analyte such as myoglobin that has clinically useful concentration increases for ~8 h after a myocardial infarction.
The antibody transduction system gave a reversible response to myoglobin binding in solution phase. This observation means that the antibody–myoglobin dissociation reaction is relatively fast (21). However, the sensor configuration did not give a reversible fluorescence response to decreased myoglobin concentration after saturation of antibody binding sites. The lack of sensor reversibility may be a result of restricted movement of dissociated myoglobin within the gel matrix, preventing diffusion away from the antibody before rebinding. Lack of sensor reversibility could limit the utility for continuous monitoring of increases and decreases in concentration. However, a sensor that can respond to increases in concentration over time may have clinical value for measuring an analyte such as myoglobin that temporally increases after release from damaged cardiac tissue. Applications of this design for in vitro measurements are not limited by reversibility considerations.
Further development of the sensor is necessary for realistic measurements in a blood or serum matrix. Optimization of the polyacrylamide gel may achieve a faster response time and also permit adjustment of the dynamic range. Different immobilization strategies could make antibody more assessable to improve response time and allow a reversible response. Optical system enhancement by use of a laser source and pulsed operation could reduce background signal and enhance the fluorescence. Use of other fluorophores may also contribute to improvements in detection limits. Higher-affinity antibodies could be used but would negatively impact reversibility.
This fiber-optic immunosensor design demonstrates a miniature self-contained system capable of real-time protein analyte measurements. The limit of detection of the current configuration was ~5 nmol/L myoglobin. The sensor had a response time generally in the range of 20–60 min. Improvements in both response time and dynamic range of concentration may be realized by further development of chemical, immobilization, and optical variables. The sensor design has barrier properties that may be suitable for whole-blood measurements.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), 675-fold (
), and 1350-fold
(
).
), D (
), F (•), and G (
). All
antibodies were labeled with 823-fold molar excess Cascade Blue.
