4.2. Antibody Bead Array Assay

Antibodies against proteins analyzed in this study were selected from the antibody repository of the Human Protein Atlas (HPA) project. A total of 259 antibodies against 130 unique proteins were selected based on concentration (>0.05 mg/mL) and binding specificity. Each antibody was coupled to a specific population of magnetic, color-coded beads (MagPlex, Luminex corp. Austin, TX, US). Bead identities coupled to rabbit anti-albumin antibodies (Dako, Abcam, Cambridge, United Kingdom) and donkey anti-human IgG antibodies (Jackson Immuno Research Laboratories Inc., Philadelphia, PA, USA) were added as positive controls. IgG from non-immunized rabbits and a bead population processed without antibodies were added as negative controls. The beads were processed, assessed and suspension bead arrays (SBAs) were generated as previously described [67].

A 25 µL aliquot of the samples from sensitive and non-sensitive breast cancer and head-and-neck cancer patients were randomized in 96-well microtiter plates prior to analysis using a liquid handling device (Freedom Evo150, Tecan, Männedorf, Switzerland). This was done to avoid bias caused by sequential instrument read-out [68]. Each plate was designed to have a similar distribution of samples concerning age and gender. Samples from the same patient were generally assigned to the same plate. In each 96-well plate, three wells were reserved for a replicated sample pool and at least three wells were kept sample-free to serve as technical controls. Three µL of the samples were then diluted 1:10 in Phosphate Buffer Saline (PBS) (Medicago, Thermo Fisher Scientific, Waltham, MA, USA) and NHS-PEO4-biotin (Pierce, ThermoFisher Scientific, Waltham, MA, USA) was used for labelling [21]. For the transfer of small volumes (<5 µL), a suitable liquid handling device (CyBi-SELMA, CyBio, Analytik Jena, Jena, Germany) was used. The labelled samples were diluted and incubated over-night at room temperature together with the SBA [21]. The beads were washed with 1xPBS 1% Tween20 using a plate washer (EL406, BioTek, Winooski, VT, USA) and an instrument containing a flow cytometer (Flexmap3D, Luminex Corp. Austin, TX, USA) was used for the fluorescence read-out. The assay was performed twice on different days to ensure the robustness and reproducibility of the procedure (1st and 2nd run).

To evaluate the selectivity of the antibodies targeting the two proteins STIM1 and THPO, a duplex sandwich immunoassay was built, as previously described [22]. To find matching antibody pairs for the assay, 3 antibodies from the Human Protein Atlas project (HPA011018, HPA011088, HPA012123) and 2 commercial antibodies (S6197 and S6072, Sigma-Aldrich, St. Louis, MO, USA) targeting STIM1, and 6 HPA antibodies (HPA019596, HPA030603, HPA042965, HPA048828, HPA051629, HPA076834) targeting THPO were coupled to a distinct bead population and were combined to form SBAs. The same antibodies were biotinylated as previously described [69], and along with one already biotinylated commercial detection antibody targeting THPO (BAF288, LOT: AKZ0415061, R&D, Systems, Bio-techne, Minneapolis, MN, USA), they were evaluated using a dilution series of plasma samples to assess the performance and duplex capacity. The built sandwich duplex-immunoassay was later conducted on 71 of the HNC patient samples. The samples were diluted 1:10 in an assay buffer consisting of 1x PBS with 0.5% (w/v) polyvinyl alcohol (P8136, Sigma-Aldrich, St. Louis, MO, USA) 0.8% (w/v) polyvinylpyrrolidone (PVP360, Sigma-Aldrich, St. Louis, MO, USA), 0.1% casein (C5890, Sigma-Aldrich, St. Louis, MO, USA) and supplemented with 0.5 mg/mL rabbit IgG (Bethyl Laboratories, Montgomery, TX, US), and incubated with the capture antibody SBA over-night on a shaker and at room temperature. The beads were then washed with 1xPBS 1% Tween20 using a plate washer (EL406, BioTek, Winooski, VT, USA), followed by incubation for 1.5 h with 25 µL of 1 µg/mL detection antibody. The beads were subsequently washed, and 50 µL of 0.5 µg/mL R-phycoerythrin-labeled streptavidin (Invitrogen, ThermoFisher Scientific, Waltham, MA, USA) in 1xPBS 1% Tween20 was added and incubated for 20 min. Lastly, the beads were washed and measured in 60 µL 1xPBS 1% Tween20 using a flow cytometer (Flexmap3D, Luminex Corp. Austin, TX, USA). Raw mean fluorescent intensity MFI values were normalized by subtracting negative control signals from bare beads without any antibody.