The Case for Human Serum as a Highly Preferable Sample Matrix for Detection of Anthrax Toxins

Abstract

This paper describes preliminary results on the surprising impact of human serum as a sample matrix on the detectability of protective antigen (PA) and lethal factor (LF), two antigenic protein markers of Bacillus anthracis, in a heterogeneous immunometric assay. Two sample matrices were examined: human serum and physiological buffer. Human serum is used as a specimen in the diagnostic testing of potentially infected individuals. Physiological buffers are often applied to the recovery of biomarkers dispersed in suspicious white powders and other suspect specimens and as a serum diluent to combat contributions to the measured test response from nonspecific adsorption. The results of these experiments using a sandwich immunoassay read out by surface-enhanced Raman scattering yielded estimates for the limit of detection (LOD) for both markers when using spiked human serum that were remarkably lower than those of spiked physiological buffer (∼70,000× for PA and ∼25,000× for LF). The difference in LODs is attributed to a degradation in the effectiveness of the capture and/or labeling steps in the immunoassay due to the known propensity for both proteins to denature in buffer. These findings indicate that the use of physiological buffer for serum dilution or recovery from a powdered matrix is counter to the low-level detection of these two antigenic proteins. The potential implications of these results with respect to the ability to detect markers of other pathogenic agents are briefly discussed.