2.2.1.

Antibody-dye conjugation

A mixture of antihuman IL-6 “detection” antibodies (0.5 mg/ml) and an optimum concentration (a dye-to-protein molar ratio of 30) of Alexa Flour 488 was prepared in PBS containing 0.1 M sodium bicarbonate (pH 8.3). The reaction mixture was incubated for 2 h at room temperature with constant shaking. The labeled antibody conjugates were separated from residual dye by use of a gel-filtration spin column (CS-800). The degree of labeling (dye-to-protein ratio) of the labeled antibody was 14. A solution of labeled “detection” anti-IL-6 antibody (1 μg/ml) was prepared with dilution buffer containing PBS and 1 mg/ml of EA.

2.2.2.

CTFOB dip-probe preparation

Quantitative measurement required identical fabrication of the probes. Each probe was an multimode optical silica/silica fiber, 8 cm long (Polymicro Technologies, Phoenix, Arizona) with a core diameter of 600 μm. Approximately 1.5 cm of protective polyimide buffer surrounding the fiber was removed from one end by burning it off with a Bunsen burner. The fiber was then decontaminated by sonication in a soap solution. This was followed by sonication in deionized water to remove any soot on the surface of the fiber. The cladding of the probe was removed by immersing the 1.5-cm uncoated part in 10% hydrofluoric acid. Some acid ascended by capillarity into the space between fiber probe and polyimide buffer. Capillary action tapered the section between the etched probe and cladded fiber. CTFOB dip probes with core diameters of 300 μm were obtained by adjusting the time of immersion. After being removed from the hydrofluoric acid, the probes were sonicated for 5 min in water, 10 min in 1 M NaOH, and for 5 min in water, and 2 min in acetone. To verify that the fabricated probes were identical, the taper angle for each probe was measured by use of a microscope. The average tapered angle for the probes was 0.058±0.002 rad. Appropriate antibodies were then immobilized on the surface of the probe.

2.2.3.

Antibody immobilization

During antibody immobilization, the orientation of attached antibodies is controlled to avoid the problem of mixed avidity because an antibody-binding site oriented toward the probe surface may cause it to be unavailable for binding its intended target. To avoid mixed avidity, the immunoglobulin (IgG antibody) is attached to the probe surface from its fab portion. Furthermore, the fluorescence signal in the CTFOB dip probe is produced by evanescent field, which decays exponentially with distance from the probe surface. Therefore, efficient signal generation requires the smallest possible distance between dye molecules and the probe surface. In view of these factors, a single cross-linker was used to immobilize IgG on the probe surface.¹⁶

Sulfhydryl residues were introduced into the hinge region of the immunoglobulin (IgG antibody) for conjugation with the maleimide-activated probe surface.²³ The cleaned probes were kept for 1 min in 2% APTS solution in dry acetone. APTS was used to derivatize the fiber surface with primary amines. These amines subsequently reacted with the hetero-bifunctional cross-linker, sulfo-SMCC, resulting in a maleimide-activated probe surface capable of reacting with sulfhydryl groups on antibodies. A sulfo-SMCC solution (4.5 mM) was prepared in PBS-EDTA (10 mM EDTA). The probes were placed into the solution for 60 min and then cleaned with the PBS-EDTA solution. To expose the sulfhydryl group of the antihuman IL-6, the capture antibodies were reduced by mixing 0.1 mg/ml of antibody solution in PBS-EDTA with MEA (50 mM). The mixture was incubated for 90 min at 37°C. After incubation, the reduced antibody molecules were desalted by use of a gel-filtration spin column (CS-800). The reduced capture antibody extract was diluted in PBS-EDTA to a concentration of 10 μg/ml. Maleimide-activated probes were incubated in reduced antibody solution for 4 h at room temperature. After being removed from the capture antibody solution, the probes were washed with PBS-EDTA.

2.2.4.

Serum samples

The two serum samples used were part of a larger study for biomarker discovery. Because these investigations have involved the serum/plasma samples collected from human subjects, in accordance with the Helsinki Declaration, we followed the Health Insurance Portability and Accountability Act Privacy and Security Rules for the research, obtained the informed consent from the subjects, and this research was approved by the institutional review board (IRB Protocol No. X090224006) of the University of Alabama at Birmingham.

Analysis was performed on small aliquots of samples collected for routine laboratory tests. For the current study, serum samples were obtained from a patient with lupus and from a patient with lymphoma before they were treated with chemotherapy. Handling and processing was similar for the samples. In brief, blood samples were collected without anticoagulant into red-top Vacutainers and allowed to coagulate for 20–30 min at room temperature. Sera were separated by centrifugation, and the specimens were immediately divided into portions, frozen, and stored in a dedicated −80°C freezer. No more than two freeze-thaw cycles were allowed before testing of samples.

2.2.5.

Standard sample preparations

Recombinant human IL-6 was used as a standard for a calibration curve. To avoid changes of sample concentrations due to adsorption on the container walls, a dilution buffer containing PBS and EA (1 mg/ml) was used to prepare IL-6 solutions at various standard concentrations (5, 10, 20, 50, 75, 100, 150, 200, 500, and 1000 pM).

2.2.6.

Signal detection setup

A schematic of the setup for signal detection is shown in Fig.1. A laser diode (Nichia, Tokyo, Japan) of 476 nm was used as an excitation source. Output from laser diode was collimated by mounting it in a collimation tube (Thorlabs Inc., Newton, New Jersey). To block any red-tail emission from the diode, a dichroic bandpass filter (Edmund Optics, Barrington, New Jeresy) with peak transmission at 476 nm and bandwidth of 10 nm was placed in front of the laser. The laser tube was mounted at one port of the collection chamber, and the CTFOB dip probe was mounted on a diagonally opposite port by use of a SMA-adaptable, bare-fiber terminator (Thorlabs Inc., Newton, New Jersey). This arrangement allowed convenient changing of the dip probes for different experiments. One end of the fluorescence collection fiber was mounted on a collection chamber port at a right angle to the laser port and dip-probe port; the other end was mounted on to the collection port of a miniature CCD-based fiber-optic spectrometer (model HR2000, Ocean Optics Inc., Dunedin, Florida). A short-focal-length focusing/collection lens (FCL) was used to focus the laser beam onto a 600-μm core dip probe. The evanescent wave-induced fluorescence from the probe surface coupled back into the probe fiber. The same FCL lens acted as a lens for fluorescence collection and collimation. A dichroic short-pass filter (Maier Photonics Inc., Manchester, Vermont) was placed in the laser beam path, making a 45-deg angle with the laser beam.

Fig. 1

Block diagram of the experimental setup: LD, laser diode; BF, bandpass filter; ND, neutral density; SF, short-pass filter; LF, long-pass filter; FL and FCL.

Because the peak emission wavelength for the Alexa 488 dye was 530 nm, a short-pass filter with high reflectivity (95%) at wavelengths of >488 nm and high transmission (90%) for the excitation wavelength at 476 nm was chosen. The short-pass filter allowed the laser light to pass but reflected the collected fluorescence toward the collection port. A focusing lens (FL) focused the fluorescence into the collection fiber. A 488-nm laser line long-pass filter (Edmund Optics, Barrington, New Jersey) was placed in front of the collection fiber. This filter has 99% transmission for wavelengths of >500 nm, but small transmission (0.0001%) for wavelengths of 500 nm. This arrangement improved the signal-to-noise ratio by blocking scattered light from the diode laser. The spectrometer was interfaced to a notebook computer, which collected the spectra.

2.2.7.

Signal generation and recording

Probes with immobilized “capture” antibodies were incubated in sample solutions for an appropriate time at room temperature. After removal from the sample (serum or standard) solution, probes were washed for 2 min in a buffer containing PBS and 0.02% Tween-20. After drying the probes, the background signal spectrum was recorded. The CCD-based spectrometer was used to record signal spectra.

The probes were incubated for an appropriate time at room temperature in labeled “detection” anti-IL-6 antibody solution (1 μg/ml). Probes were washed for 2 min, then dried before recording the signal spectral profile. The probes were washed for an additional 2 min and dried before the signal was rerecorded. If a probe showed no signal change, then there was no further washing. Otherwise, 2-min washings and subsequent signal recordings continued until no change in signal was observed. Typically, no more than two 2-min washings were required.

2.2.8.

Signal extraction by least-squares fitting method

The signal spectrum profile is a mixture of two profiles: an autofluorescence (background) profile f _B(λ) from nonspecific bindings and the characteristic fluorescence profile f _Fl(λ) of labeled “detection” antibodies. Mathematically, the recorded signal spectral profile f _Signal(λ) is a linear combination of the two:

A standard fluorescence spectral profile (entire spectrum) f _Fl(λ) of Alexa 488 was recorded after immobilizing only the labeled “detection” antibodies on the probe surface. The background spectral profile (entire spectrum) f _B(λ) for each probe was recorded before incubating the probe in labeled detection antibodies. After this incubation, the signal spectral profile (entire spectrum) f_Signal(λ) for each probe was recorded.

Least-squares fitting to obtain coefficients a and b was achieved by use of the “solver method” of Microsoft Excel. The sum shown in Eq. 2 was obtained from the data of the three recorded profiles, signal f_Signal(λ), background f_B(λ), and standard profile f_Fl(λ). Initially guessed values of a and b were used,

By minimizing the sum shown in Eq. 2 by use of a “solver method,” actual values of coefficients a and b were obtained. The ratio of coefficients, b/a, for each probe was taken as the normalized value of the fluorescence signal of that probe.

2.2.9.

Bead-based xMAP technology

xMAP technology (Luminex Corp., Austin, Texas) combines the principles of a sandwich immunoassay with fluorescent beads, allowing individual and multiplex analysis of up to 100 different analytes in a single microtiter well. The technique involves polystyrene microspheres internally dyed with differing amounts of two spectrally distinct fluorophores to create a family of 100 different, spectrally addressed sets of beads. Each bead set can be conjugated with a capture antibody specific for a unique target protein. In a multiplexed assay, antibody-conjugated beads are allowed to react with sample (serum or standard). After washing of the beads, detection antibodies are added to a microtiter plate well to form a “capture sandwich” immunoassay. The assay solution is analyzed by a fluorometric array reader, which obtains two fluorescence readings for each bead: one that identifies a bead as a member of one of the 100 sets and another that measures the amount of fluorescent dye, phycoerythrin (PE), bound to the detection antibody. The amount of green fluorescence is proportional to the amount of analyte captured in the immunoassay. Bio-Plex Manager software correlates each bead set to the assay reagent coupled to it. Extrapolating to a standard curve allows quantitation of each analyte in the sample.

The xMAP assay for human cytokines IL-2, IL-4, IL-6, IL-8, IL-10, IFN-γ, TNF-α, and granulocyte macrophage colony-stimulating factor (Human Cytokine 8-Plex Assay, Bio-Rad, Hercules, California) was accomplished in 96-well microplates according to the Bio-Rad protocol, as previously described. Samples were analyzed by use of the Bio-Plex suspension array kit (Bio-Rad Laboratories, Hercules, California). Beads, immobilizing by purified antigen (by the vendor), were incubated with sera diluted 1:4, and quantitation was accomplished by incubating the beads with PE-conjugated rabbit antihuman Abs (IgG). The minimum cytokine detection level for these kits was <5 pg/ml. The values were presented as mean ± SE (pg/mL). For these studies, only the xMAP values for IL-6 were utilized to compare the estimates of IL-6 obtained by CTFOB.