Endogenous peroxidases in sputum interfere with horse-radish peroxidase-based ELISAs

Peroxidase-based immunoassays are commonly used for detecting inflammatory mediators in biological samples. We suggest caution while interpreting assays particularly in sputum samples that have endogenous peroxidases like eosinophil peroxidase (EPX), which may interact with a horseradish peroxidase (HRP)-based ELISA. Using IL-8 as an example, we demonstrate that values generated with an HRP-ELISA (n=47) show significant positive correlation with the sputum EPX content (r=0.6, P=0.0004), which can be misconstrued to be affiliated with an eosinophilic event. The data-set generated with the same samples (n=47) using alkaline phosphatase (AP)-based ELISA and a non-enzymatic Milliplex system do not show any correlation with sputum EPX (Milliplex r=-0.24, P=0.13; AP r=0.26, P=0.09). Moreover, sub-group analysis shows significantly increased IL-8 levels detected by HRP-ELISA in eosinophilic patient sputa (n=28) compared to AP-ELISA (P=0.0001). We, therefore, recommend the use of AP-based ELISA or Multiplex system rather than peroxidase-based ELISA for detecting soluble mediators, and more importantly for non-Th2 related mediators in sputum samples with increased eosinophil activity.