Validation and comparison of ELISA kits to measure interferon gamma responses in QuantiFERON cultural supernatants for diagnosis of tuberculosis

Use of Polymer Micropillar Arrays as Templates for Solid-Phase Immunoassays

We investigate the use of periodic micropillar arrays produced by high-fidelity microfabrication with cyclic olefin polymers for solid-phase immunoassays. These three-dimensional (3D) templates offer higher surface-to-volume ratios than two-dimensional substrates, making it possible to attach more antibodies and so increase the signal obtained by the assay. Micropillar arrays also provide the capacity to induce wicking, which is used to distribute and confine antibodies on the surface with spatial control. Micropillar array substrates are modified by using oxygen plasma treatment, followed by grafting of (3-aminopropyl)triethoxysilane for binding proteins covalently using glutaraldehyde as a cross-linker. The relationship between microstructure and fluorescence signal was investigated through variation of pitch (10-50 μm), pillar diameter (5-40 μm), and pillar height (5-57 μm). Our findings suggest that signal intensity scales proportionally with the 3D surface area available for performing solid-phase immunoassays. A linear relationship between fluorescence intensity and microscale structure can be maintained even when the aspect ratio and pillar density both become very high, opening the possibility of tuning assay response by design such that desired signal intensity is obtained over a wide dynamic range compatible with different assays, analyte concentrations, and readout instruments. We demonstrate the versatility of the approach by performing the most common immunoassay formats-direct, indirect, and sandwich-in a qualitative fashion by using colorimetric and fluorescence-based detection for a number of clinically relevant protein markers, such as tumor necrosis factor alpha, interferon gamma (IFN-γ), and spike protein of severe acute respiratory syndrome coronavirus 2. We also show quantitative detection of IFN-γ in serum using a fluorescence-based sandwich immunoassay and calibrated samples with spike-in concentrations ranging from 50 pg/mL to 5 μg/mL, yielding an estimated limit of detection of ∼1 pg/mL for arrays with high micropillar density (11561 per mm2) and aspect ratio (1:11.35).