True Sensitivity of Immunoassays: Are Concentrations of Low Abundance Analytes Real or Artifacts?

We developed immunoassays with fg/mL detection limits, 100 to 1,000X more sensitive than current ELISA technology. We measured concentrations of low-abundance cytokines and other biomarkers in the serum of healthy donors, and found that these analytes are present at lower concentrations than had been reported in previous studies. This raised the question of how to demonstrate the reliability of immunoassay measurements for analytes present at sub-pg/mL concentrations. To establish whether measured concentrations are due to the true analyte or an artifact, we depleted the analyte from clinical samples by using high quality antibodies conjugated to Dynabeads. We hypothesized that high abundance serum or plasma components that bind non-specifically to immunoassay components would not be depleted, and that the true analyte would be depleted by the antibodies depending on affinity. IL-2, IL-4, IL-6, IL-10, IL-17A, IL-21, GM-CSF, TSLP, TNF-a, Cardiac Troponin I, and Tau were used as model analytes. The limits of detection achieved for these S-PLEX™ assays ranged from ~0.2 to ~20 fg/mL. Spike recovery and dilution linearity were between 80% – 120%. Depletion experiments were performed with up to six antibodies per assay. All samples could be depleted by high quality antibodies, including antibodies that were not part of the assays. This indicates that the assays measured the true analyte. In conclusion, using depletion experiments, we demonstrated that our assays are highly specific and measure the correct analytes. This test of “True Sensitivity” can be applied to all assays claiming to measure low abundance biomarkers.