Comparison of Immunochemical (Enzyme-Linked Immunosorbent Assay) and Immunohistochemical Methods for the Detection of Central Nervous System Tissue in Meat Products

Abstract

Three methods are widely used in the United States to detect the presence of central nervous system (CNS) tissue in meat products: the fluorescent enzyme-linked immunosorbent assay (F-ELISA), developed in this laboratory, the colorimetric Ridascreen Risk Material 10/5 ELISA (R-ELISA), and the U.S. Department of Agriculture, Food Safety and Inspection Service immunohistochemical (IHC) procedure. These assays are based on the immunological detection of glial fibrillary acidic protein (GFAP), a neural antigen largely restricted to the CNS. The objective of the current study was to compare the sensitivity and repeatability of these tests for detecting the presence of neural tissue in meat. Ground beef spiked with 0.05 to 0.5% of bovine brain, spinal cord (SC), or dorsal root ganglia, as well as advanced meat recovery samples, were evaluated by each of the three GFAP detection procedures. Interassay coefficients of variation for the F-ELISA GFAP standards were 7 to 25%, and intra-assay variation due to sampling and extraction of spiked ground beef was 7 to 13% for SC and 8 to 14% for brain (n = 10). The F-ELISA was the most sensitive of the methods tested, capable of detecting 0.3 ng GFAP standard per well and the presence of 0.05% brain and SC in meat. The R-ELISA standards produced highly variable results (up to 36% variation) and, as a result, none of these standards were different from zero (n = 26). The R-ELISA resulted in high sample variation in SC-spiked ground beef samples (coefficients of variation were 23 to 50%) and did not detect the presence of brain contamination. After modification of the R-ELISA sampling and extraction methods, SC-spiked sample variation was reduced to 16 to 20%, and sensitivity was improved from 0.3 to 0.2% SC, although brain tissue was still not detected. The IHC analysis of CNS-adulterated ground beef had a sensitivity of 0.2% SC and 0.05% brain, with false-negative rates of 10 to 20% at and above the stated sensitivities. None of the methods examined detected dorsal root ganglia contamination. The F-ELISA detected the presence of CNS contamination in 20% of the advanced meat recovery samples, compared to 3.5 to 5% for the R-ELISA and 2% for IHC. This study suggests that the F-ELISA is much more sensitive and repeatable than either the R-ELISA or the IHC procedure method for the detection of CNS tissue in meat products.