Parallel comparison of sandwich and direct label assay protocols on cytokine detection protein arrays

This study describes a systematic comparison of the preparation, sensitivity, and response of direct label and sandwich fluoroimmunoassay protein arrays for parallel detection of cytoldnes and growth factors. Five model analytes were examined, IL-1beta, TNF-alpha, VEGF, MIP-1beta, and TGF-beta1, as well as four different array printing slides. Each slide was printed with four identical anti-cytokine arrays, two arrays employed for direct labeling and two for sandwich assays. All five target cytokines assayed using the sandwich format significantly outperformed the corresponding direct label assays in terms of background-subtracted fluorescent intensity, although the extent varied for different cytokines, growth factors, and slides examined. Binding of target protein by immobilized capture antibody was examined by Biacore surface plasmon resonance with and without labeling of the target protein. A major explanation for the superior performance of the sandwich assay protocol was the dilution of the labeled target cytokine in the size exclusion gel filtration column during labeling procedure. Other possible explanations included that binding of the detection antibody-induced signal amplification, and labeling of target cytokine reduced efficiency of cytokine binding to immobilized capture antibody.