Electrochemical immunomagnetic assay for interleukin-6 detection in human plasma.

Conceptualization, G. B., O. E. O., H. A. R., and D. E. C.; methodology, G. B. and O. E. O.; validation, G. B.; formal analysis, G. B.; investigation, G. B.; writing–original draft preparation, G. B.; writing–review and editing, G. B., O. E. O., H. A. R., and D. E. C.; visualization, G. B.; supervision, G. B. and D. E. C.; project administration, G. B. and D. E. C.; funding acquisition, D. E. C.; resources, D. E. C.

IL-6 antibodies were conjugated to epoxy-coated magnetic beads using the manufacturers protocol and buffers (C1, C2, HB, LB, and SB). Epoxy-coated magnetic beads were washed with 1 mL of C1 buffer, mixed, placed on magnet, and the supernatant was removed. 149 μL of C1 buffer, 0.9 μL of 100 mg mL −1 IL-6 antibody, and 150 μL of C2 buffer were combined, mixed with beads, and allowed to incubate at 37 °C for 16–24 h while mixing. After mixing, the solution was placed on magnet and the supernatant was removed. 800 μL of HB buffer was added, mixed, placed on magnet, and supernatant was removed. This was repeated with LB buffer once, and SB buffer four times. Then 800 μL of SB buffer was mixed at room temperature for 15 minutes before placing on magnet and removing supernatant. The beads were resuspended in 300 μL SB to create a 10 mg mL −1 stock and stored at 4 °C until use. We found that aliquoting the beads into 10 μL aliquots and stored ready to be used in assays improved the replicability of the assays. Beads were typically only stored for up to one month before use; however, the product manual cites stability in solution for at least one year. 10 μL of magnetic bead stock were used per assay and mixed with 50 μL of IL-6 and 2 μL of biotinylated detection antibody per assay. In order to prevent agglomeration, the biotinylated detection antibody was diluted in buffer containing 0.1% Tween-20. Then, the mixture was placed on mixer at room temperature for 30 minutes. After mixing, the mixture was placed on magnet stand and the supernatant was removed with a transfer pipette after one minute. Three washes were performed to reduce nonspecific binding but also to ensure minimal bead loss. The washes were performed by resuspending beads in 1 mL of 1× PBS 0.01% Tween-20, placing on magnet stand for one minute, rinsing out pipette tip, waiting one minute, and removing supernatant with transfer pipette. This was done twice with 1× PBS with 0.01% Tween-20. One wash was done with PBS that did not contain Tween-20. After the third wash, 15 μL of 4 μg mL −1 streptavidin poly-HRP was added to the beads and the mixture was placed on the mixer for 5 minutes. The poly-HRP enzyme was conjugated to the detection antibody via streptavidin-biotin coupling. After streptavidin polyHRP incubation, the mixture was placed on the magnet stand and supernatant was removed with transfer pipette after one minute. The three washes were repeated as before, with two washes with 1× PBS with 0.01% Tween-20 and one wash with PBS containing no Tween-20. Finally, 100 μL of TMB was added to the beads, mixed several times until beads are resuspended, and placed on mixer at room temperature for 5 min. The mixture was removed from mixer with 2 min remaining, then placed on magnet stand with 1.5 min remaining, the supernatant was removed at 0.5 min remaining and dropcasted onto a bared unactivated CSPE. At 0 minutes remaining, electrochemical reduction was performed using chronoamperometry for 20 seconds at −0.1 V. We integrated the current to collect total charge passed in 20 seconds to plot with the IL-6 calibrant concentration. IL-6 solutions were measured in PBS, DMEM, and human plasma. Many measurements were made in PBS to optimize the system and verify its repeatability. Repeatability for DMEM and human plasma were tested at least three times each.

Scheme 1 displays the electrochemical immunoassay steps based on using magnetic beads and polymeric HRP. Chronoamperometry was used to detect TMB diimine reduction. Then, the current of the chronoamperometry curve was integrated to get the total charge, which is plotted with the calibrant concentration. In chronoamperometry, the current in the initial milliseconds is largely controlled by charging current. Since nonfaradaic current affects the later parts of the current-time curve less, using the total charge passed produces better signal-to-noise ratios. Furthermore, chronocoulometric data is cleaner as integration smooths the random noise from the current-time curves. As a result, we report charge vs. concentration for IL-6 calibrations. We first verified replicability of the TMB reduction using chronoamperometry. In order to verify that the reduction of oxidized TMB was sufficiently replicable across different CSPEs, assay size was increased by 6 times and incubated with 50 ng mL −1 IL-6-HRP then split six ways before incubation with TMB. The six TMB solutions were electrochemically reduced using chronoamperometry ( Fig. 1 ). These assays demonstrated sufficient replicability, with a percent difference of 8.3% between the largest and smallest charges integrated over 20 seconds. Thus, it was determined that the HRP/TMB enzyme/substrate system was appropriate for moving forward with assay optimization and calibration. Next, we calibrated the sensor in PBS. Two calibrations from 0.05–1 ng mL −1 IL-6 in PBS were run on two separate days ( Fig. 2a ). They exhibited good day to day replicability, with the largest percent difference of 11.0% occurring at 0.75 ng mL −1 . These calibrations also demonstrated excellent linearity, with an R 2 of 0.988. In order to verify the performance of the assay in complex media, we tested three calibration replicates from 0.05–1 ng mL −1 IL-6 in DMEM cellular media ( Fig. 2b ). One replicate was run on one day and the other two were run on another. They exhibited desirable day to day replicability, with the largest percent difference of 10.6%. This calibration also exhibited excellent linearity, with an R 2 of 0.993. Next, we ran triplicate calibrations in plasma ( Fig. 2c ) which showed an increase in variability, as expected. The sensitivity of this calibration was much less compared to calibrations in other media. We suspect that this is due to a combination of the batch of detection antibody and the matrix. Future work includes applying the IL-6 sandwich immunoassay as a potential diagnostic assay in a wide variety of biological samples.

The polyHRP/TMB system combined with magnetic beads for IL-6 capture allowed for the detection of IL-6 in the physiologically relevant concentration range. The sandwich assay demonstrated excellent linearity and replicability and achieved detection in 50–1000 pg mL −1 range. Overall, the assay exhibited good performance in biologically complex solutions including cellular media and human plasma and shortened the time required for IL-6 concentration readout. Replicability, linearity, and assay time were all improved compared to similar previously published works. Additionally, the simplicity of the assay procedure further improves upon previously reported designs. Future work involves validating the assay's measurements with conventional ELISAs and applying this immunoassay to a wide variety of biological systems as a potential diagnostic.

Recombinant Human Interleukin-6 (IL-6, M w = 23 718 g mol −1 ), was purchased (Recombinant Human, Carrier Free) from R&D Systems. Recombinant Human Matrix Metalloproteinase-9 (MMP-9, M w = 93 kDa) and Recombinant Human Matrix Metalloproteinase-3 (MMP-3, M w = 54 kDa) were purchased from Sigma Aldrich, USA. KCl (Certified ACS), MgCl2, NaCl, Trizma base, 10× phosphate buffer solution, IL-6 Human matched antibody pair (CHC1263), Dynabeads antibody coupling kit (14311D), and Dulbecco's Modified Eagle Media (DMEM) were obtained from Thermo Fisher Scientific. The following chemicals were used as received without additional purification and were obtained from Sigma Aldrich (St. Louis, MO, USA): 3,3′,5,5′-tetramethylbenzidine (TMB) liquid substrate system for ELISA (T0440). Normal patient pooled plasma was obtained from Jonathan Schoenecker's lab and spiked with IL-6 for assay calibration in human plasma. All experiments were performed in accordance with the NIH and Vanderbilt IRB Guidelines, and approved by the ethics committee at Vanderbilt University. De-identified human plasma was obtained with informed consent from human donors. Electrochemical measurements were performed using a CHI 1440 4-Channel Potentiostat (CH Instruments, Austin, TX). Carbon screen-printed electrodes (CSPE) from Pine Research Instrumentation, Inc. with a 2 mm diameter carbon working electrode, a Ag/AgCl reference electrode, and a carbon counter electrode were used for electrochemical interrogation. All experiments were completed on a bare unactivated CSPE.

The authors declare no conflicts of interest.