Figure 3.

Robustness of the sandwich immunoassays in various complex matrices. Dilutions of target toxins, i.e., recombinant lab-made SEA at 250 pg/mL (a), SEG at 50 pg/mL (b), SEH at 50 pg/mL (c), and SEI at 333 pg/mL (d), were performed in EIA buffer (positive control reaching approximately 1 optical density (OD) signal), undiluted matrices (black bars), 2-fold diluted matrices in EIA buffer (dark gray bars), 5-fold diluted matrices in EIA buffer (clear gray bars), or 20-fold diluted matrices in EIA buffer (white bars). Undiluted liquid matrices consisted of raw matrices buffered by the addition of a one-tenth volume of 10X EIA buffer. Undiluted solid matrices consisted of 10% (W/V) ground homogenate prepared in water and buffered by the addition of a one-tenth volume of 10X EIA buffer. All these spiked samples were detected using the best-identified sandwich immunoassay (Table 3) using the 3-h sequential format with poly-horseradish peroxidase-labeled streptavidin detection. Results are expressed as a percentage of signal (optical density with subtraction of nonspecific binding) obtained when the target toxin is prepared in EIA buffer (black bar on the right side of each graph). Numbers located above the bars indicate the theoretical LoD (calculated as explained in methods) measured in the corresponding matrix and calculated from one experiment. Error bars represent standard deviations from two independent experiments. N.C., not calculable. N.D., not determined.