Protein microarrays are an increasingly popular tool for detecting proteins and antibodies as well as probing their interactions. Investigators usually detect interactions on protein arrays using fluorescently labeled secondary antibodies followed by direct comparison of signal intensities, allowing for qualitative assessment but limiting the ability for quantitative assessment. Lower throughput methods such as ELISA commonly take advantage of a set of binding standards so that signal intensities may be quantified by comparison with a standard curve. Previous attempts have been made to use standard curves with protein arrays, but the linear curve fitting approaches that were used limited the antibody binding detection range to one order of magnitude, which is inappropriate for concentrations at the high and low ends of the curve. Yu et al. sought to solve this problem through the use of microarray nonlinear calibration (MiNC) which they describe in this issue of BioTechniques. The authors’ protein array was generated with a nucleic acid programmable protein array where cDNAs are printed on the array and then converted to proteins by in vitro transcription and translation. The resulting proteins adhere to a capture agent on the array via a corresponding epitope tag. MiNC arrays also carried a series of known quantities of IgG standards. After probing arrays with antibodies or serum expected to interact with one of the protein targets along with the appropriate secondary antibody coupled to a fluorescent dye, arrays were imaged and signals detected. The authors next fitted a nonlinear curve to the signals from the labeled secondary antibody binding the IgG standards and compared this with the target feature signals to quantify the amount of antibody that bound the surface of each spot. This approach provided a linear relationship for a range of 2.5-3.0 orders of magnitude and decreased the signal variation both within and between experiments. With careful fitting of the nonlinear curve, MiNC enables the quantification of binding changes, making the approach particularly useful for biomarker discovery and clinical diagnostics.
Blood platelets respond to multiple stimuli in a donor specific manner to initiate blood clotting. Interactions among multiple clotting signaling pathways can result in synergistic or antagonistic effects if they use common second messengers. While the prothrombotic state of a platelet is determined by its integrated response to various activating and inhibiting signals, previous in vitro studies have only examined the response of platelets to single agonists. A greater understanding of the interactions of different platelet signaling pathways requires examination of co-stimulation by agonists since phosphatidylserine exposure occurs with pairs of agonists but not in response to a single agonist. In order to analyze the effects of pairs of agonists added simultaneously to platelets, D. Jaeger and S. Diamond at the University of Pennsylvania (Philadelphia, PA) took advantage of their multiplexed pairwise agonist-flow cytometry (PAS-FC) method to measure platelet inside-out responses to pairwise combinations of agonists. Six agonists (convulxin, SFLRRN, AYPGKF, ADP, U46619, and PGE2) that stimulate platelet signaling through the GPVI, PAR-1, PAR-4, P2Y1, P2Y12, TP, and IP receptors were used as a representative selection of the major signals received by a platelet during thrombosis. Three-color flow cytometry of the platelets was used to examine platelet response by simultaneous measurement of integrin αIIbβ3 activation with PAC-1 antibody, P-selectin exposure (induced by alpha granule release) with anti-P-selectin antibody, and phosphatidylserine exposure with annexin V. Using 96-well microplates, platelet-rich plasma samples were mixed with fluorescently labeled PAC-1, anti-P-selectin, and annexin V and then with all 15 possible pairwise combinations of the 6 agonists, after which flow cytometry with automated well plate handling was carried out. The authors analyzed platelets from 10 healthy donors using the PAS-FC method and found that for duplicate measurements, 4 out of the 10 donors had sufficiently unique 45-parameter (15 agonist pairs × 3 colors) phenotypes to self-cluster. With further refinements, such as using multiple agonist concentrations, the assay will allow efficient analysis of donor-specific responses to various agonists.