Development of a sensitive ELISA for the quantification of human tumour necrosis factor-alpha using 4 polyclonal antibodies.

Despite the availability of many assays to measure concentrations of tumour necrosis factor alpha (TNF-alpha) in body fluids, these assays often lack specificity or sensitivity and are often of questionable reliability, resulting in inconsistent results. Therefore, we have developed an ELISA that is sensitive, reliable and not susceptible to disturbances by interfering substances such as heterophilic antibodies. The assay involves a combination of four polyclonal antibodies. The antibodies, which capture the analyte, were raised in chicken and the trapping anti-analyte antibodies were raised in rabbit. The immobilization of capture antibodies was achieved via a coating antibody raised in a duck against chicken IgY and the recognition of trapping antibodies was achieved by a detection antibody raised in a goat against rabbit IgG and labelled with HRP. The analytical and functional sensitivities of the ELISA are 8 pg/mL and 13 pg/mL, respectively. The assay showed good precision and, in contrast to our in-house RIA, excellent parallelism in serial dilutions. The recovery of TNF-alpha spiked to plasma samples ranged from 97% to 119%. Comparison of the newly developed, sensitive ELISA with our in-house RIA showed that the median TNF-alpha value obtained by RIA (range: 0.095-10.0, median 0.578 ng/mL) was found to be 1.5-2 times higher than that obtained with the ELISA (range 0.008-5.84, median 0.213 ng/mL). Spearman correlation was 0.755 (p < 0.0001). In addition, analysis of the TNF-alpha concentrations in blood from healthy individuals and from patients suffering from tuberculosis, with RIA and ELISA, showed the same differences although TNF-alpha levels obtained with ELISA were lower. We feel that this ELISA is a major improvement compared to the currently available assays for TNF.