Development of Multiplexed Bead-Based Immunoassays for Profiling Soluble Cytokines and CD163 Using Mass Cytometry

Abstract

Bead-based immunoassays are multiparametric analysis allowing for the simultaneous quantification of a large number of biomarkers within a single sample. Mass cytometry is an emerging cytometric technique that offers a high multiplexing capacity in a high-throughput setting but has not yet been applied to bead-based assays. In this study, we developed a multiplex bead-based immunoassay of cytokines and CD163 designed for mass cytometry (MC). A set of 11 types of lanthanide-encoded microbeads were synthesized by two-stage dispersion polymerization as classifier candidates for the assay. These beads were then decorated with different Abs on the surface to capture the target cytokines in solution. Gold nanoparticles were employed as reporters to identify the binding of target cytokines on the classifier surface. As a proof-of-concept study, we first developed four-plex and nine-plex assays of mixtures of cytokines in standard solutions. The MC signal intensities of these immunoassays were responsive to the concentration differences in the standard solutions with high detection sensitivities at low analyte concentrations. Finally, we examined a sample of peripheral blood mononuclear cells (PBMCs) with the nine-plex assay, comparing an unstimulated sample with a sample stimulated to promote cytokine secretion.

Scheme 1. Schematic Diagram for the Multiplexed Bead-Based Sandwich Immunoassays by MC Carried out in a 96-Well Filter Plate

Metal-encoded microbeads were surface modified with capture Abs and employed as the classifier to recognize the target analytes in samples. The detailed protocol is described in the Experimental Section.

Figure 1

(a) ¹⁴⁰Ce–¹⁴²Ce isotopic dot-plot diagram of a mixture of 11 types of classifier microbeads (C-1 to C-11). The oval circle isolates the singlet events of 11 types of microbeads. (b–k) Dot-plot diagrams showing the gating strategies to individually identify C-1 to C-11 microbeads by MC.

Figure 2

Histograms of the reporter signal intensities on IL-4 classifier beads (C-5) in a series of four-plex assays of standard solutions at various IL-4 concentrations. (a–c) AuNP was employed as the reporter in four-plex assays of standard solutions containing IL-4 at concentrations of 0, 1.2, and 20 pg/mL, respectively. (d–f) NanoGold was employed as the reporter in four-plex assays of standard solutions containing IL-4 at 0, 1.2, and 20 pg/mL, respectively.

Figure 3

Standard curves of two sets of four-plex assays for (a) IL-4, (b) IL-6, (c) IFNγ, and (d) TNFα. The x-axis in each plot represents the analyte concentration, and the y-axis represents the median MC signal intensity of NPs attached to the corresponding classifier beads. Two different types of streptavidin-conjugated reporter (AuNP and NanoGold) were investigated in these four-plex assays. The results are presented for the AuNP as blue-filled circles (●) and for the NanoGold as green filled squares (■). Negative events with ¹⁹⁷Au signal intensities of ≤1 count per bead were excluded from the statistical analysis for median intensities. The dose–response curves were drawn by fitting the experimental results with a four-parameter logistic regression model.

Figure 4

Standard curves of four sets of nine-plex assays for (a) IL-1β, (b) IL-4, (c) IL-6, (d) IL-10, (e) IL-18, (f) IFNγ (g) TNFα, (h) CD163, and (i) CXCL-9. The x-axis in each plot represents the analyte concentration. The y-axis in each plot represents the median MC signal intensity of AuNP attached to the corresponding type of classifier beads. To minimize the background noise at low analyte concentrations, the concentrations of biotinylated anti-CD163 and anti-CXCL9 in the detection Ab cocktails were reduced from 2.5 to 2.0, 1.0, and 0.5 μg/mL, while the concentrations of other detection Abs were kept constant at 2.5 μg/mL in the cocktails. The results of these assays are plotted for anti-CD163 and anti-CXCL9 Ab concentrations of 2.5 μg/mL with filled circles (●), of 2.0 μg/mL with filled squares (■), of 1.0 μg/mL with filled triangles (▲), and of 0.5 μg/mL with filled diamonds (◆). Negative events with ¹⁹⁷Au signal intensities of ≤1 count per bead were excluded from the statistical analysis for median intensities. The dose–response curves were drawn by fitting the experimental results with a four-parameter logistic regression model.

Figure 5

Median ¹⁹⁷Au signal intensities of AuNP reporter attached to classifier beads in nine-plex assays for the analysis of (a) IL-1β, (b) IL-4, (c) IL-6, (d) IL-10, (e) IL-18, (f) IFNγ, (g) TNFα, (h) CD163, and (i) CXCL-9 in the stimulated and unstimulated PBMC samples at different sample dilution ratios. To measure the extracellular cytokine release by PBMCs, the stimulated and unstimulated samples were collected from the supernatants of the PBMC suspensions in culture media with/without PMA/ionomycin stimulation. Solid columns in the figure represent the assay results of the stimulated samples, while striped columns represent the assay results of the unstimulated samples. In a typical assay, sample solution (50 μL) without dilution or diluted at the ratios of 2×, 8×, 32×, or 128× was added to the classifier dispersion (50 μL). The detection Ab concentrations of anti-CD163 and anti-CXCL9 in the assays were 0.5 and 1.0 μg/mL, while the concentration of other detection Abs was 2.5 μg/mL.