Naturally occurring carbohydrate antibodies: Interference in solid-phase immunoassays

Antibody assays utilizing carbohydrate matrices (agarose, cellulose) for protein insolubilization are subject to non-specific and specific interfering factors. This report examines factors which diminish the quality of particle-based solid-phase radioimmunoassays (SPRIAs) for antigen-specific human IgG. Interfering factors are divided into (a) constant non-specific binding which is similar with all human sera and appears related to simple adherence of IgG to agarose and cellulose, and (b) absorbable binding which varies considerably among sera in agarose and cellulose-based assays, and results from the presence of IgG antibody specific for the carbohydrate matrix. Constant non-specific binding is predictably 1-3% B_max (maximum binding) in all human and rabbit sera. In contrast, the absorbable binding levels vary widely: 0.75-29% B_max in 58% (study 1, n = 50) and 40% (study 2, n = 200) of normal individuals for agarose, and 3-30% B_max in 70% of the population for microcrystalline cellulose. IgG anti-agarose antibodies were found in 13 of 16 rabbit sera examined. Ultracentrifugation and immune-complex studies demonstrated that aggregated or immune-complexed IgG does not contribute to the absorbable IgG binding. Inhibition with acid hydrolyzed soluble agarose and provided evidence for a specific IgG anti-agarose antibody that causes variable background binding. Pre-absorption of sera with agarose prior to analysis in the agarose-based SPRIA removed > 90% of anti-agarose antibodies and eliminated false positive results. These studies suggest rabbit and perhaps other heterologous antibodies prepared by protein-agarose affinity column chromatography may contain significant levels of naturally occurring antibodies against agarose or cellulose. These naturally occuring carbohydrate antibodies may interfere in solid-phase carbohydrate-based immunologic methods and immunoassays.