Development of a non-competitive immunoassay for cortisol and its application to the analysis of saliva

A general method of performing non-competitive immunoassays for a low-molecular-mass analyte was developed and applied to cortisol determination in saliva samples. The method is based on the use of a “blocking reagent”, which is able to bind to antibody sites not occupied by the analyte, and in a stronger way than the analyte itself. When an enzyme-labelled analyte is added it substitutes the analyte in the antibody complex, but not the blocking reagent. The measured signal is linearly correlated to the concentration of the complex and, consequently, to the analyte concentration. The 3σ limit of detection (LOD, 0.2 nM) obtained by the above method was 10 times lower than that obtained by the corresponding ELISA. As non-competitive immunoassays reported for small molecules up to now have been no more than just approaches, the suitability of the proposed assay for cortisol quantification in a real matrix was investigated. Human saliva was chosen as a matrix because of the need for very sensitive techniques to determine salivary cortisol content. The matrix effect was offset by performing the calibration experiments in acidic conditions (pH = 5.6) and adding 0.1% of bovine serum albumin (BSA) to the buffer. In these conditions, the LOD was 1.4 nM, which was adequate to measure normal levels of cortisol. Spiked samples were analysed and gave recoveries ranging from about 80 to 120%. Therefore, five subject samples, collected over 18 h showed salivary cortisol concentrations compatible with the circadian variation of reported normal values